Strategic Applications of Distance Learning Technologies (Advances in Distance Education Technologies) (Premier Reference Source)

Strategic Applications of Distance Learning Technologies Mahbubur Rahman Syed Minnesota State University, Mankato, USA

InformatIon scIence reference Hershey • New York

Director of Editorial Content: Managing Development Editor: Senior Managing Editor: Managing Editor: Assistant Managing Editors: Copy Editor: Typesetter: Cover Design: Printed at:

Kristin Klinger Kristin M. Roth Jennifer Neidig Jamie Snavely Carole Coulson Alana Bubnis Carole Coulson Lisa Tosheff Yurchak Printing Inc.

Published in the United States of America by Information Science Reference (an imprint of IGI Global) 701 E. Chocolate Avenue, Suite 200 Hershey PA 17033 Tel: 717-533-8845 Fax: 717-533-8661 E-mail: [email protected] Web site: http://www.igi-global.com and in the United Kingdom by Information Science Reference (an imprint of IGI Global) 3 Henrietta Street Covent Garden London WC2E 8LU Tel: 44 20 7240 0856 Fax: 44 20 7379 0609 Web site: http://www.eurospanbookstore.com Copyright © 2009 by IGI Global. All rights reserved. No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher. Product or company names used in this set are for identification purposes only. Inclusion of the names of the products or companies does ot indicate a claim of ownership by IGI Global of the trademark or registered trademark. Library of Congress Cataloging-in-Publication Data

Strategic applications of distance learning technologies / Mahbubur Rahman Syed, editor. p. cm. Summary: "This collection of advanced research incorporates global challenges and opportunities of technology integration while outlining strategies for distance learning within developing countries"--Provided by publisher. Includes bibliographical references and index. ISBN 978-1-59904-480-4 (hardcover) -- ISBN 978-1-59904-482-8 1. Distance education--Computer-assisted instruction. 2. Educational technology. I. Syed, Mahbubur Rahman, 1952LC5803.C65S77 2008 371.35'8--dc22 2008007624 British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. All work contributed to this book set is original material. The views expressed in this book are those of the authors, but not necessarily of the publisher.

If a library purchased a print copy of this publication, please go to http://www.igi-global.com/agreement for information on activating the library's complimentary electronic access to this publication.

Editorial Advisory Board

Patricia Lipetzky Minnesota State University, Mankato, USA Orlando Baiocchi University of Washington, Tacoma, USA Marco Roccetti University of Bologna, Italy Timothy K. Shih Tamkang University, Taiwan Jianhua Ma Hosei University, Tokyo, Japan Fernando Gamboa-Rodríguez Universidad Nacional Autónoma de México, Mexico

Table of Contents

Preface ................................................................................................................................. xiv Acknowledgment ................................................................................................................ xvii

Chapter I Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS ...............................1 Sérgio Deusdado, Instituto Politécnico de Bragança, Portugal Paulo Carvalho, Universidade do Minho, Portugal Chapter II Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach: A Longitudinal Field Experiment ....................................................................14 Charlie C. Chen, Appalachian State University, USA R. S. Shaw, Tamkang University, Taiwan Chapter III Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment .......................................................................................................................30 Charlie C. Chen, Appalachian State University, USA Albert L. Harris, Appalachian State University, USA Rong-An Shang, Soochow University, Taiwan Chapter IV Motivation-to-E-Learn: A Quantitative Design Technique ................................................................49 M. A. Rentroia-Bonito, Technical University of Lisbon, Portugal J. A. Jorge, Technical University of Lisbon, Portugal C. Ghaoui, John Moores University, UK Chapter V Algorithm Education Using Structured Hypermedia.........................................................................58 Tomasz Müldner, Acadia University, Canada Elhadi Shakshuki, Acadia University, Canada Andreas Kerren, Växjö University, Sweden

Chapter VI Federated Agent-Based Architecture for Collaborative Education Model ........................................84 Iwona Miliszewska, Victoria University, Australia Chapter VII An Agent-Based Framework for Personalized Learning in Continuing Professional Development .................................................................................................................96 Apple W. P. Fok, City University of Hong Kong, Hong Kong Horace H. S. Ip, City University of Hong Kong, Hong Kong Chapter VIII Development and Evaluation of a Keyword-Accessible Lecture Video Player and Lecture Video Contents....................................................................................................................111 Takahiro Yoshida , Tokyo University of Science, Japan Seiichiro Hangai, Tokyo University of Science, Japan Chapter IX Distance Learning in Business Aviation Industry: Lessons Learned and Implications for Theory and Practice ....................................................................................................................124 Mahesh S. Raisinghani, TWU School of Management, USA Chris Colquitt, National Aeronautics and Space Administration, USA Mohammed Chowdhury, University of Dallas, USA Chapter X SEAMAN: A Visual Language-Based Tool for E-Learning Processes ...........................................147 Gennaro Costagliola, University of Salerno, Italy Filomena Ferrucci, University of Salerno, Italy Giuseppe Polese, University of Salerno, Italy Giuseppe Scanniello, University of Salerno, Italy Chapter XI An Architecture for Online Laboratory E-Learning System ...........................................................165 Bing Duan, Nanyang Technological University, Singapore Habib Mir M. Hosseini, Nanyang Technological University, Singapore Keck Voon Ling, Nanyang Technological University, Singapore Robert Kheng Leng Gay, Nanyang Technological University, Singapore Chapter XII A Virtual Laboratory for Digital Signal Processing ........................................................................ 180 Chyi-Ren Dow, Feng Chia University, Taiwan Yi-Hsung Li, Feng Chia University, Taiwan Jin-Yu Bai, Feng Chia University, Taiwan

Chapter XIII Information Retrieval in Virtual Universities ..................................................................................194 Juha Puustjärvi, Helsinki University of Technology, Finland Päivi Pöyry, Helsinki University of Technology, Finland Chapter XIV A Web-Based Tutor for Java™: Evidence of Meaningful Learning ................................................207 Henry H. Emurian, University of Maryland, Baltimore County, USA Chapter XV Personalisation in Web-Based Learning Environments ...................................................................230 Mohammad Issack Santally , University of Mauritius, Mauritius Senteni Alain, University of Mauritius, Mauritius Chapter XVI Implementation and Performance Evaluation of WWW Conference System for Supporting Remote Mental Health Care Education.........................................................................251 Kaoru Sugita, Fukuoka Institute of Technology (FIT), Japan Giuseppe De Marco, Fukuoka Institute of Technology (FIT), Japan Leonard Barolli, Fukuoka Institute of Technology (FIT), Japan Noriki Uchida, Global Software Corporation, Japan Akihiro Miyakawa, Nanao City, Ishikawa Prefecture, Japan Chapter XVII Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning ..............................269 Degan Zhang, University of Science and Technology of Beijing, China Yuan-chao Li, China University of Petroleum, P.R. China Huaiyu Zhang, Northwest University, China Xinshang Zhang, Jidong Oilfield, P.R. China Guangping Zeng, University of Science and Technology of Beijing, China Chapter XVIII Digital Rights Management Implemented by RDF Graph Approach ..............................................284 Jin Tan Yang, Southern Taiwan University of Technology, Taiwan Huai-Chien Horng, National Kaohsiung Normal University, Taiwan

Compilation of References ............................................................................................................304 About the Contributors .................................................................................................................329 Index

............................................................................................................................................334

Detailed Table of Contents

Preface ................................................................................................................................................ xiv Acknowledgment ............................................................................................................................... xvii Chapter I Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS .................................. 1 Sergio Deusdado, Instituto Politécnico de Bragança, Portugal Paulo Carvalho, Universidade do Minho, Portugal Chapter I discusses a new generation of e-learning development, based on synchronous groupware applications integration, providing improved interactivity and pro-human relations, allows richer training experiences far beyond a virtual classroom. Despite WWW service evolution, e-conferencing multimedia applications remain “killer applications” and insensitive to resources degradation, in fact, the quality of service (QoS) provided by the network is still a limitation impairing their performance. Such applications have found in multicast technology an ally contributing for their efficient implementation and scalability. Additionally, considering QoS as design goal at application level becomes crucial for groupware development, enabling QoS proactivity to applications. Congregating these technological contributions, an adaptive platform has been developed integrating public domain multicast tools, applied to a Web-based distance learning system. The system is user-centered (e-student), aiming at good pedagogical practices and proactive usability for multimedia and network resources. The services provided, including QoS adapted interactive multimedia multicast conferences (MMC), are fully integrated and transparent to end-users. QoS adaptation, when treated systematically in tolerant real-time applications, denotes advantages in group scalability and QoS sustainability in heterogeneous and unpredictable environments such as the Internet. Chapter II Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach: A Longitudinal Field Experiment ....................................................................................... 14 Charlie C. Chen, Appalachian State University, USA R. S. Shaw, Tamkang University, Taiwan Chapter II discusses the online training and the question of whether the most effective training methods applied in live instruction will carry over to different online environments in the long run. Behavior modeling (BM) approach—teaching through demonstration—has been proven as the most effective approach

in a face-to-face (F2F) environment. A quasi-experiment was conducted with 96 undergraduate students who were taking a Microsoft SQL Server 2000 course in a university in Taiwan. The BM approach was employed in three learning environments: F2F, online synchronous and online asynchronous classes. The results were compared to see which produced the best performance, as measured by knowledge neartransfer and knowledge far-transfer effectiveness. Overall satisfaction with training was also measured. The results of the experiment indicate that during a long duration of training no significant difference in learning outcomes could be detected across the three learning environments. Chapter III Challenges in Delivering Case-Based Teaching in the Online Asychronous Learning Environment ......................................................................................................................................... 30 Charlie C. Chen, Appalachian State University, USA Albert L. Harris, Appalachian State University, USA Rong-An Shang, Soochow University, Taiwan Chapter III assesses and compares the efficacy of case method teaching in face-to-face and online asynchronous learning (OAL) environments. The overall findings of this study indicate that an online asynchronous environment can promote students’ participation in certain cases. As most antagonists for the adoption of online asynchronous case method surmised, cognitive learning gains via this learning method do not seem to be as high as in the face-to-face environment. The findings provide ample room for a further exploration of creative online asynchronous methods to continuously improve cognitive gains of learners. Chapter IV Motivation-to-E-Learn: A Quantitative Design Technique .................................................................. 49 M. A. Rentroia-Bonito, Technical University of Lisbon, Portugal J. A. Jorge, Technical University of Lisbon, Portugal C. Ghaoui, John Moores University, UK Chapter IV discusses e-learning’s challenge to promote effectiveness in order to fully get expected benefits. Achieving effectiveness will contribute to its establishment as a credible way to support educational endeavours. This work explores a variable called “motivation-to-e-learn,” a key component to design technology-supported learning experiences. Our goal is to identify what motivation-related variables are critical for student engagement in learning online. We further explored the importance of a set of motivation-to-e-learn variables building on previous results in real instructional settings. From this activity, an exploratory two-factor structure emerged which explains 96% of motivation to e-learn construct. We discuss our results, together with their implications for learning-support design and future work. Our contribution is a step towards quantitatively understanding and cost-effectively improving the link among learning-design process, supporting systems and students into an effective and harmonious whole.

Chapter V Algorithm Education Using Structured Hypermedia ........................................................................... 58 Tomasz Müldner, Acadia University, Canada Elhadi Shakshuki, Acadia University, Canada Andreas Kerren, Växjö University, Sweden Understanding of algorithms is one of the most challenging aspects of the study of computer science. In Chapter V, we present a new approach for explaining algorithms that aims to overcome various pedagogical limitations of the current visualization systems. The main idea is that, at any given time, a learner is able to focus on a single problem. This problem can be explained, studied, understood, and tested before the learner moves on to study another problem. The Structured Hypermedia Algorithm Explanation (SHALEX) system is the system we designed and implemented to explain algorithms at various levels of abstraction. Since the system is implemented using a client-server architecture, it can be used both through distance education and in the classroom setting. To aid and monitor the leaner, we also developed an agent in SHALEX that provides help and monitors the completion rate. Chapter VI Federated Agent-Based Architecture for Collaborative Model .......................................................... 84 Iwona Miliszweska, Victoria University, Australia Chapter VI presents the development of a conceptual, operational, and software architecture of a collaborative education model. The federated model, supported by agent-based communication over the Internet, can operate across geographical, cultural and organisational boundaries while promoting integration within those boundaries. Because of its potential ability to cross the various boundaries, the proposed model seems particularly applicable to distance education environments. Chapter VII An Agent-Based Framework for Personalized Learning in Continuing Professional Development.... 96 Apple W. P. Fok, City University of Hong Kong, Hong Kong Horace H. S. Ip, City University of Hong Kong, Hong Kong Chapter VII discusses the requirement of continuous professional development (CPD) activities to stay qualified for membership. Modern day professionals who are very much mobile and work within tight schedules point to the need of an asynchronous learning environment that provides a learner-centered approach and offers learners greater flexibility and choices. In this article we argue that “personalization learning” (PL) that exploits the abundance of information and e-learning materials on the Web can be harnessed effectively to serve the diversity of CPD training needs. Moreover, we specialize in the concept of PL to Personalized CPD Learning and highlight the emerging technologies that are relevant to the development of personalized learning for CPD. We further proposed an agent-based architectural and conceptual framework for a personalized CPD learning portal (Personalized-CPD) which integrates these technologies to provide supportive functions for professionals to conduct CPD activities in a personalized manner.

Chapter VIII Development and Evaluation of a Keyword-Accessible Lecture Video Player and Lecture Video Contents .............................................................................................................................................. 111 Takahiro Yoshida, Tokyo University of Science, Japan Seiichiro Hangai, Tokyo University of Science, Japan In Chapter VIIII the authors developed a lecture video player/maker system (Yoshida, 2002, 2003). In developing this system, we considered the usability for students and operability for teachers. The player includes a keyword access function, which enables the student to jump to scenes where one of the registered keywords was spoken. For this purpose, the lecture video maker realizes automatic index generation after continuous speech recognition of the whole lecture stream. Evaluations of the lecture videos and the player by students are discussed, and the desirable style of lecture videos for students is surveyed. Chapter IX Distance Learning in Business Aviation Industry: Lessons Learned and Implications for Theory and Practice ........................................................................................................................................ 124 Mahesh S. Raisinghani, TWU School of Management, USA Chris Colquitt, National Aeronautics and Space Administration, USA Mohammed Chowdhury, University of Dallas, USA Chapter IX explores the expectations and behaviors of business aviation pilots towards online learning. The authors believe that the company that is able to offer an integrated, individualized, and useful online training experience will gain a significant competitive advantage. To that end, the authors have researched and synthesized studies that are currently available and relate to this important future product. In addition, an exploratory survey of business aviation pilots and interviews with key aviation industry players are used to determine current attitudes and expectations towards online learning. The scope of this chapter will be limited to exploring the niche market of business aviation pilots using the aviation training company CAE SimuFlite and their new SimfinityTM .technology. However, the authors consider the concepts discussed to be applicable to all business aviation pilots. Chapter X SEAMAN: A Visual Language-Based Tool for E-Learning Processes ............................................. 147 Gennaro Costagliola, University of Salerno, Italy Filomena Ferrucci, University of Salerno, Italy

Giuseppe Polese,University of Salerno, Italy

Giuseppe Scanniello, University of Salerno, Italy

Chapter X concerns the design phase in the development of e-learning courses concerns. In this chapter we present a tool based on a suite of visual languages, which has been specifically conceived to support instructional designers in the definition and creation of learning processes. The proposed suite of visual languages includes the learning activity diagram, which extends UML activity diagrams to make them suitable for modelling e-learning processes, the Self-Consistent Learning Object language used to define knowledge contents, and the Test Maker Language for specifying assessment and self-assess-

ment tests. The visual languages have been then implemented in SEAMAN (System for E-Learning Activity MANagement), a system prototype conceived to support instructional designers in the design, the generation, and the deployment of e-learning processes. Chapter XI An Architecture for Online Laboratory E-Learning System .............................................................. 165 Bing Duan, Nanyang Technological University, Singapore Habib Mir M. Hosseini, Nanyang Technological University, Singapore Keck Voon Ling, Nanyang Technological University, Singapore Robert Kheng Leng Gay, Nanyang Technological University, Singapore With the goal of bringing e-learning to the traditional laboratory experiment, Chapter XI presents an architecture for an online laboratory e-learning system to facilitate the design and deployment of lab-based courses for e-education. The chapter provides an overall view of the system design and implementation so the Internet-based laboratory can be easily integrated with the e-learning infrastructure. Chapter XII A Virtual Laboratory for Digital Signal Processing ........................................................................... 180 Chyi-Ren Dow, Feng Chia University, Taiwan Yi-Hsung Li, Feng Chia University, Taiwan Jin-Yu Bai, Feng Chia University, Taiwan In Chapter XII the authors design and implement a virtual digital signal processing laboratory (VDSPL). VDSPL consists of four parts: mobile agent execution environments, mobile agents, DSP development software, and DSP experimental platforms. The network capability of VDSPL is created by using mobile agent and wrapper techniques without modifying the source code of the original programs. VDSPL provides human-human and human-computer interaction for students and teachers, and it also can lighten the teacher’s load, increase the learning result of students, and improve the usage of network bandwidth. A prototype of VDSPL has been implemented by using the IBM Aglet system and Java Native Interface for DSP experimental platforms. Also, experimental results demonstrate that our system has received many positive feedbacks from both students and teachers. Chapter XIII Information Retrieval in Virtual Universities .................................................................................... 194 Juha Puustjärvi, Helsinki University of Technology, Finland Päivi Pöyry, Helsinki University of Technology, Finland Chapter XII discusses information retrieval in the context of virtual universities which deals with the representation, organization, and access to learning objects. In this chapter, we give an overview of the ONES system, and analyze the relevance of two information retrieval models for virtual universities. We argue that keywords based search (i.e., the Boolean model), though well suited for Web searches, is overly coarse for virtual universities. Instead, the vector model, on which our implemented search engine is also based on, seems to be more appropriate as it provides similarity measure (i.e., the learning object having the best match is presented first). We also compare the performance of four algorithms for computing the similarities (matching).

Chapter XIV A Web-Based Tutor for Java™: Evidence of Meaningful Learning ................................................... 207 Henry H. Emurian, University of Maryland, Baltimore County, USA In Chapter XIV students in a graduate class and an undergraduate class in Information Systems completed a Web-based programmed instruction tutor that taught a simple Java applet as the first technical training exercise in a computer programming course. The tutor is a competency-based instructional system for individualized distance learning with the capacity to generate meaningful learning (i.e., understanding of concepts) at the level of the individual student. Chapter XV Personalisation in Web-Based Learning Environments ...................................................................... 230 Mohammad Issack Sentally, University of Mauritius, Mauritius Senteni Alain, University of Mauritius, Mauritius Chapter XV proposes a framework for research in promoting personalisation in Web-based learning environments. The concepts of adaptability, adaptivity and the limitations of completely adaptive systems are discussed. The conception of more interactive environments that are both adaptable and adaptive, which can assist the teacher in making interesting pedagogical decisions while tutoring in a virtual environment is proposed. Two versions of an algorithm that can be used to offer personalisation in the framework described are developed and discussed in this chapter. The algorithm is basically a method devised to select the most appropriate learning object from a pool of potential objects that exist in the repository. Chapter XVI Implementation and Performance Evaluation of WWW Conference System for Supporting Remote Mental Health Care Education ............................................................................................. 251 Kaoru Sugita, Fukuoka Institute of Technology (FIT), Japan Giuseppe DeMarco, Fukuoka Institute of Technology (FIT), Japan Leonard Barolli, Fukuoka Institute of Technology (FIT), Japan Noriki Uchida, Global Software Corporation, Japan Akihiro Miyakawa, Nanao City, Ishikawa Prefecture, Japan Information technology (IT) can be helpful for remote mental health care education. Because there are very few mental health care specialists, it is very important to decrease their moving time. But it is not easy to use the conventional TV conference systems for ordinary people, mental health care specialists, and their students because they are not computer specialists. For this reason, we have developed a WWW conference system. Our system can communicate between the mental health care specialists and their students by using the live video on WWW browser. In this paper, we show the implementation and the evaluation of proposed system. The experimental results over the Internet show that our system can be used for real time communication between Fukuoka, Ishikawa, and Iwate prefectures.

Chapter XVII Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning ................................ 269 Degan Zhang, University of Science and Technology of Beijing, China Yuan-chao Li, China University of Petroleum, P.R. China Huaiyu Zhang, Northwest University, China Xinshang Zhang, Jidong Oilfield, P.R. China Guanping Zeng, University of Science and Technology of Beijing, China In Chapter XVII, under the banner of seamless mobility, we propose a kind of approach supporting taskoriented mobile distance learning paradigm. Web-based seamless migration, which has the capability that task for mobile distance learning (MDL) dynamically follows the learner from place to place and machine to machine without learner’s awareness or intervention by active service. Our key idea is this capability can be achieved by architecture of component smart platform and agent-based migrating mechanism. Chapter XVIII Digital Rights Management Imlemented by RDF Graph Approach ................................................... 284 Jin Tan Yang, Southern Taiwan University of Technology, Taiwan Huai-Chien Horng, National Kaohsiung Normal University, Taiwan Chapter XVIII proposes a design framework for constructing digital rights management (DRM) that enables learning objects in legal usage. The central theme of this framework is that any design of a DRM must have theories as foundations to make the maintenance, extension or interoperability easy. Two algorithms for encoding and verifying rights in DRM are designed to deal with REL metadata in RDF format. This technological support also reduces the sophistication among role assignments, learning objects and task ontology of DRM. The DRM module is embedded to SCORM-compliant content repository management system (CRMS) for IPR (intellectual property rights) protection.

Compilation of References .............................................................................................................. 304 About the Contributors ................................................................................................................... 329 Index ................................................................................................................................................... 334

xiv

Preface

EmErging TEchnologiEs in DisTancE lEarning It is an established fact that for a successful distance education program we require technologies that provide increased interactivity between the learners and teachers. It is just about a decade since the understanding of technology for distance education were meant to include post, telephone, fax and limited use of the Internet. The rapid growth in computers, telecommunication technologies and capabilities, and change in nature of the information age has significantly changed the definition and technology requirements in distance education. Technologies considered as advanced for distance education in the recent past are abandoned and new technologies are demanded today. The advances in technology have caused paradigm shifts in distance education starting from correspondence courses, teleconferencing over speaker phones, teleconferencing via modem, transporting still pictures along with interactive audio, to the latest technology of two-way, full audio, full video communication. Current technological advances have generated a great deal of excitement and hope to overcome the walls and boundaries, the barriers of real time interactivity in distance learning education. Today development of learning modules that include elements such as video transmission, e-mail, the Internet, and the World Wide Web supported by multimedia are common. The goal is to minimize or overcome the limits of separation between the learners, educators and facilitators by time and distance. This book focuses chapters on different research, design and implementation aspects of technologies and methods with specific focus on distance education. These include: •

•

•

•

Development of integrated e-learning environment based on interactive multimedia services with proactive QoS; allowing development of end-to-end QoS-aware multimedia conferences, coordinating resources from network, end-system processing equipment and applications. One proposed system integrates public domain multicast applications for synchronous media communication, being supervised by a middleware-based QoS management framework intending to preserve the QoS of critical parameters for e-learning session's specificity. Studies to demonstrate how one of the most effective training methods, the behavior modeling (BM) approach, that is, teaching through demonstration, applied in live instruction will carry over to three different online environments: F2F, online synchronous and online asynchronous classes. Assessment and comparison between the efficacy of case method teaching in face-to-face and online asynchronous learning (OAL) environments. Four hypotheses are proposed on the correlation between these two delivery modes and studied the learning performance of students. Research on “motivation-to-e-learn,” a key component to design technology-supported learning experiences, with focus on quantitative approaches to support learning-centered design by consid-

xv

•

•

•

• • • •

• •

•

•

• •

•

•

ering student needs and their immediate and broader contexts to promote effectiveness in order to fully get the expected benefits of e-learning challenges. New approach for explaining algorithms that aims to overcome various pedagogical limitations of the current visualization systems through design and implementation of Structured Hypermedia Algorithm Explanation (SHALEX) system to explain algorithms at various levels of abstraction. Presentation of a collaborative education model that would provide efficient communication services and an open scalable architecture for the uniform publication, management, and dissemination of distributed educational material developed by geographically dispersed educational providers, while maintaining the autonomy of the participating providers. Design of an agent-based architectural and conceptual framework for a Personalized Continuous Professional Development Learning Portal (Personalized-CPD) that, by harnessing the abundance of information and e-learning materials on the Web, can be effectively used to serve the diversity of CPD training needs. Discussion and development of a keyword-accessible lecture video player to enable students to view past lectures at any time and from anywhere on their PCs. Provide understanding of the expectations and behaviors of business aviation pilots towards online learning. Introduction of a tool based on a suite of visual languages, which has been specifically conceived to support instructional designers in the definition and creation of learning processes. Design and implementation of different Internet-based virtual laboratories, a rapidly growing research area in universities, to facilitate the designing and deployment of the lab-based courses for e-education. Implementation of a prototype of a virtual digital signal processing laboratory (VDSPL) by using the IBM Aglet system and Java Native Interface for DSP experimental platforms. Research on information retrieval in the context of virtual universities and dealing with the representation, organization, and access to learning objects. The representation and organization of learning objects should provide the learner with an easy access to the learning objects. One of the challenges in developing an automated distance learning system, which is to craft the instructional experience so that students acquire the capability to solve problems not explicitly taught or encountered in the system itself. Introduction to a series of formative evaluations to assess and enhance the instructional effectiveness of an automated and individualized distance learning system that is intended to assist information systems students in beginning their study of Java™. Investigation of the problem of personalization in Web-based learning environments. Development and evaluation of a WWW conference system in order to realize a remote mental health care education by providing communication between the mental health care specialists and their students in addition to providing communication between the mental health care specialists, patients and their families by using the live video on WWW browser, point-to-point communication, point-to-multipoint communication and multipoint-to-multipoint communication. Development of a kind of approach supporting a task-oriented mobile distance learning paradigm—Web-based seamless migration, which has the capability that task for mobile distance learning (MDL) dynamically follows the learner from place to place and machine to machine without learner’s awareness or intervention by active service, which may be achieved by architecture of component smart platform and agent-based migrating mechanism. A design framework recommendation for constructing digital rights management (DRM) that enables learning objects in legal usage. The central theme of this framework is that any design of a

xvi

DRM must have theories as foundations to make the maintenance, extension, or inter-operability easy. The chapters in this book reinforce the fact that the digital revolution, powered by the engines of information and communication technologies, has fundamentally changed the way people think, behave, communicate, work and earn their living. It has restructured the means by which the world conducts economic and business activities and runs governments. It has formed new ways to create knowledge, educate people and disseminate information.

xvii

Acknowledgment

Many people deserve credit for successful publication of this book. I express my sincere gratitude to each of the chapter authors in this book, who contributed and expanded all the ideas mentioned above and made their expertise available in bringing this book to fruition. Support from colleagues and staff in the Department of Information Systems and Technology and the administration at Minnesota State University Mankato helped sustain my continued interest. A special note of thanks goes to all staff at IGI Global, whose contribution throughout the whole process from inception of the initial idea to final publication has been invaluable. I am grateful to my wife Sharifun and my son Tahin who by their unconditional love have steered me to this point and given me constant support. Mahbubur Rahman Syed Editor

xviii

About the Editor

Mahbubur Rahman Syed is currently a professor of information systems and technology at Minnesota State University, Mankato (MSU), USA. He has about 25 years of experience in teaching, in industry, in research and in academic leadership in the field of computer science, engineering, information technology and systems. Earlier he worked in the Electrical and Computer Engineering Department at the North Dakota State University in USA, in the School of Computing and Information Technology, Monash University in Australia, in the Department of Computer Science and Engineering in Bangladesh University of Engineering and Technology (BUET) in Bangladesh and Ganz Electric Works in Hungary. He was a founding member of the Department of Computer Science and Engineering at BUET and served as Head of the Department during 1986-92. He served as the general secretary of Bangladesh Computer Society and also as the General Secretary of BUET Teacher's Association. He received the UNESCO/ROSTSCA' 85 award for South and Central Asia region in the field of Informatics and Computer Applications in Scientific Research. He won several other awards. He has co-edited several books in the area of e-commerce, software agents, multimedia systems and networking. He guest edited the 2001 fall issue of IEEE Multimedia. He has more than 100 papers published in journals and conference proceedings. He has been serving in different roles such as co-editor-in chief, associate editor, editorial review committee, member of several international journals. Dr. Syed has been involved in international professional activities including organizing conferences and serving as conference and program committee chair.



Chapter I

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS Sérgio Deusdado Instituto Politécnico de Bragança, Portugal Paulo Carvalho Universidade do Minho, Portugal

absTracT A new generation of e-learning development, based on synchronous groupware applications integration, providing improved interactivity and pro-human relations, allows richer training experiences far beyond a virtual classroom. Despite WWW service evolution, e-conferencing multimedia applications remain “killer applications” and insensitive to resources degradation, in fact, the quality of service (QoS) provided by the network is still a limitation impairing their performance. Such applications have found in multicast technology an ally contributing for their efficient implementation and scalability. Additionally, considering QoS as design goal at application level becomes crucial for groupware development, enabling QoS proactivity to applications. The applications’ ability to adapt themselves dynamically according to the resources availability can be considered a quality factor. Tolerant real-time applications, such as videoconferences, are in the frontline to benefit from QoS adaptation. However, not all include adaptive technology is able to provide both end-system and network quality awareness. Adaptation, in these cases, can be achieved by introducing a multiplatform middleware layer responsible for tutoring the applications’ resources (enabling adjudication or limitation) based on the available processing and networking capabilities. Congregating these technological contributions, an adaptive platform has been developed integrating public domain multicast tools, applied to a Web-based distance learning system. The system is user-centered (e-student), aiming at good pedagogical practices and proactive usability for multimedia and network resources. The services provided, including QoS adapted interactive multimedia Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

multicast conferences (MMC), are fully integrated and transparent to end-users. QoS adaptation, when treated systematically in tolerant real-time applications, denotes advantages in group scalability and QoS sustainability in heterogeneous and unpredictable environments such as the Internet.

inTroDUcTion Technology has been a strong catalyst for educational innovation and improvement, especially when the World Wide Web is involved. The next generation Internet needs technological support to accommodate promising new applications, such as interactive real-time multimedia distribution. Next generation e-learning platforms will support cooperative use of geographically distributed educational resources as an aggregated environment, thus enabling a more effective knowledge exchange but facing several challenges, such as flexibility, extensibility, and scalability (Amoretti, Bertolazzi, Reggiani, Zanichelli, & Conte, 2005). Predictable bandwidth availability and capacity solvency imply QoS management to regulate resources in heterogeneous environments. IP multicasting techniques (Deering, 1998; Kosiur, 1998; Moshin, Wong, & Bhutt, 2001; Thaler & Handley, 2000; Ratnasamy, Ermolinskiy, & Shenker, 2006) are attractive solutions for the capacity shortage problem, as bandwidth consumption is reduced when network resources are shared. On the other hand, the QoS support (Moshin, Wong, & Bhutt, 2001) should be, in a first instance, inherent to applications in order to integrate conveniently enhanced real-time multimedia applications in the present Internet, barely QoS aware and increasingly heterogeneous. With the advent of wireless and mobile networks, heterogeneity is likely to subsist; so envisioned applications should merge QoS adaptation and multicast in a proactive utilization of resources. Applications should be designed with adaptation in mind; they need to employ built-in mechanisms that allow them to probe the condi-



tions of the network environment and alter their transmission characteristics accordingly (Miras, 2002). Self-adaptive applications, in the sense of proactive behavior for transmission of continuous media in multiparty applications, are a well-accepted solution due to the correct integration of new services in today’s Internet (Lubonski, Gay, & Simmonds, 2005; Deusdado, 2002; Li, Xu, Naharstedt, & Liu, 1998). E-learning, as a component of flexible learning, encompasses a wide set of applications and processes that use available electronic media to deliver vocational education and training. It includes computer-based learning, Web-based learning, virtual classrooms and digital collaboration (Eklund, Kay, & Lunch, 2003). Our work aims to integrate interactive multimedia e-learning applications in a proactive fashion taking into account the available network resources and QoS sustainability. In this way, our motivation is to offer improved learning experience based on ultimate technology with QoS warranties. The system architecture proposed in this work includes an adaptive module based on Java applets and embedded Javascript, responsible for assessing the existing operating conditions by collecting metrics reflecting the client’s end-system performance (e-student’s host), the current network conditions and relevant multicast group characteristics. The collected data is subsequently computed weighting parameters such as the available bandwidth at the client side, the round-trip time between the client and the e-learning server, the client’s current CPU load and free memory. The obtained results are used for proper multicast applications scheduling and parameterization in a transparent way.

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

multiway communications (point-to-multipoint and multipoint-to-multipoint), satisfying most of the needs of group communication, such as e-learning services. Using these applications by common e-students drives recurrently to poor QoS satisfaction due to the heterogeneity of resource conditions and the applications’ inability to assess available conditions and adjust internal parameters before conference initiation. Without regulation, real-time transport protocol (RTP) traffic floods the network capacity insensitively, forcing network congestion in certain cases or inhibiting better performance. A coherent behavior of an application without adaptation is difficult in today’s Internet. Public domain multicast applications used in this work vic (McCanne & Jacobson, 1995), rat (Hardman, Kirstein, Sasse, Handley, & Watson, 1995) and Java Media Framework (JMF) (JMF 2.0, 1999) were designed with no QoS “sensors,” so the communication dynamics are not automatically interdependent of end-systems or network conditions. Effectively, such applications allow

moTiVaTions Basically, e-learning services are used to promote connections between people (e-students) and training resources (Steeples & Jones, 2002). E-learning research is wide and growing in importance, especially in higher education. Several institutions are developing interactive Web-based learning systems, integrating rich media streaming, which may compromise network performance. The design of e-learning systems should consider QoS as mandatory for successful learning experience, selecting the appropriate technologies and applications, and regulating proactively the information and communications technology (ICT) resources utilization (Lubonski, Gay, & Simmonds, 2005; Allison, Ruddle, McKechan, & Michaelson, 2001). The Multicast Backbone (MBone) is a network overlaying the global Internet designed to support multipoint applications. MBone tools comprise a collection of audio, video and whiteboard applications that use Internet multicast protocols to enable

Figure 1. QoS tolerance for generic audio and video applications (Miras, 2002) packet loss desired

acceptable

audio/video streaming -%

interactive video interactive audio

-%

Figure 1 – QoS tolerance for generic audio and video applications (Miras, 2002). 3 |

|

|

0ms 00ms 00ms

| 00ms

| -0seg

delay



Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

pre-parameterization to adjust critical parameters such as throughput, number of frames per second, video and audio encoding formats, and so on. Adaptation, in these cases, can be obtained by introducing a multi-platform middleware layer responsible for tutoring the applications’ resources (adjudication or limitation) based on the available processing and networking capabilities (Miras, 2002). Common interactive real-time applications are fault-tolerant but suffer from QoS constraints; lowlatency requirements, and reliability are cumulative to achieve conference success. The diagram in Figure 1 attempts to illustrate the QoS tolerance, in terms of delay and packet loss, for generic interactive audio and video applications. The main motivation of this work is to provide adaptive behavior to applications used on both sides of multimedia conferencing, focusing essentially on multicast members that initiate audio and/or video transmission. The underlying idea is to launch automatically MMC applications with proper audio and video codecs, bandwidth allocation inference, and other parameters that affect sustainability and scalability during an elearning session. Our emphasis is on the concept of “interactive e-learning services,” relegating the concept of “e-learning course” to a secondary goal, which will be considered in future work. Most prominent related work on friendly multimedia transmission over the Internet, based on a combination of system and network QoS feedback, implementing equation-based adaptation is summarized in Bouras and Gkamas (2003) and Vandalore, Feng, Jain, & Fahmy (2001).

sYsTEm’s archiTEcTUrE For multicast video distribution to heterogeneous users in an e-learning session, we assume that a class server (e-tutor’s system) should be distributed and platform independent, considering inclusively multi-tutoring. Thus, a class server should con-



nect to an e-learning server (Web server) and be submitted to adaptation as a regular new sender. The QoS requirements for the class server, operating in a centralized fashion, may justify the need of layered multicast (Johanson & Lie, 2002; Liu, Li, & Zhang, 2004), enhancing the service’s adaptation. However, this work aims at integrating e-students with heterogeneous equipment when they transmit audio and video to the group, as it happens in a conventional classroom. If a client (estudent) wants to interact and multicast video, then the system’s architecture will integrate him with fair adjustments attending both to his connection to the server (e-tutor) network and to his hardware processing capabilities. Client’s adaptation should not depend on the other group members because they are transient, and consequently stability of transmissions could be very poor. As the involved applications are characterized by an intensive use of host and network resources, the purpose of the middleware platform is to achieve by computation, in a scale of five differentiated modes, the proper integration of new multicasting members. Within this thematic, it means implementing an adaptive learners’ participation in e-learning sessions by starting MMC applications transparently, with their functionality optimized for the current operating conditions. To clarify these aspects Figure 2 illustrates the system architecture. As shown, three applets, operating sequentially and interdependently, are responsible for monitoring and assessing QoS conditions; inferring, announcing and/or editing computed adaptation parameters. The process culminates with the initiation of MMC applications, depending on the host and network profiles and covering eventual end-user explicit requirements. Audio and video encoding formats, frame rate, and other quality metrics may be chosen according to the resources’ availability, providing coherent, friendly and fair participation in the network load balance. After monitoring sustainable network QoS with repeated measurements during

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

Figure 2. Proposed system’s architecture E-learning system - services Multicast multimedia conference applications Rat

Vic





Application’s QoS profile

Application’s QoS profile

middleware

applet Multicast Multicastapplications applications adaptator adaptator(audio (audioand andvideo) video)

LiveConnect HTML Forms

Client’s QoS profile

applet Parameter Parametervisualization visualization(edition (editionisis still stillpossible) possible)

LiveConnect applet

LiveConnect

Available Available resources resourcesmonitor monitor

Figure 2 – Proposed system's architecture. operating system

In the proposed framework, QoS management is performed individually for Properties and available processing resources each new conference member and occurs before the transmission's start, i.e., MMC client’s systemare launched adaptively facing the previous QoS sensing period applications the applications. If an e-student experiences lack of QoS while conferencing, his Available bandwidth and elapsed round-trip time membership process should be restarted. Corroborating this practice, MMC applications, especially multicast capable network vic, are not stable enough. In fact, if some critical adjustments are made on-the-fly, the result is often the collapse of the application. Nevertheless, Information

approximately 15 seconds before media transmission, the round-trip time (RTT) and bandwidth are calculated using a moving average. In addition, system’s status variables such as processor load, free memory, processor performance, and so on are acquired taking advantage of operating system facilities. The system is multiplatform as the included applets differentiate the most popular operative systems (Windows and Unix), invoking appropriate inner services to obtain instant measures for the processor’s load and free memory. The col-

Adaptation

lected data constitutes another input to compute an adaptation index. Different compilations were produced for common browsers. All the adaptation process is transparent, however, regarding the experimental nature of this work, each phase allows interaction with the user, providing technical information or even accepting user preferences. To achieve this goal, applets and HTML forms interchange data using Sun’s Liveconnect technology.



Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

aDaPTiVE Qos FramEWorK In the proposed framework, QoS management is performed individually for each new conference member and occurs before the transmission’s start, that is, MMC applications are launched adaptively facing the previous QoS sensing period conditions. QoS variability during the conference is not used to dynamically readapt the applications. If an e-student experiences lack of QoS while conferencing, his membership process should be restarted. Corroborating this practice, MMC applications, especially vic, are not stable enough. In fact, if some critical adjustments are made on the fly, the result is often the collapse of the application. Nevertheless, dynamic adaptation is currently subject of study within group communication applications (Layaida & Hagimonte, 2002; Tusch, Böszörményi, Goldschmidt, Hellwagner, & Schojer, 2004). Considering the applications’ specificity and type of traffic generated, adaptability only includes interactive audio (rat) and video (vic) applications and services. The heuristics regarding the choice of applications’ QoS parameters emerged from experimental results and scientific references in this matter (Wu, Hou, Zhu, Zhang, & Peha, 2001). For instance, videoconference users typically require better audio quality than video quality (Bolot, Crépin, & Garcia, 1995). The success of videoconferencing communication also depends on factors such as received frames per second, image quality, resolution, size and illumination.

For this work, the representative parameters of vic and rat used to modulate QoS are presented in Table 1. The values for these parameters, deriving from a mathematical expression that generates an adaptation mode based on the sustainable QoS level, compose a set of adjusting directives determining the applications’ behavior. Each adaptation mode indexes the respective set of adjustments, which will then be passed to the application. Since QoS scale varies from mode 1 to 5, when the obtained result is under or over this range it will be assigned to the nearest limit. Equation (1) determines the adaptation mode to be applied: M = (int) (B/(RTT/2) + FM/P) *K where1, M = QoS adaptation Mode (Table 2); B = Bandwidth (kbps); RTT = Round-Trip Time (ms); FM = Free Memory (MB); P = Processor load (%); K = 1/50 - constant to scale the result (1 to 5). For vic (version 2.8), the video encoding formats H.261 (ITU-T H.261, 1993) and H.263 (ITU-T H.263, 1998) were those that revealed best performance for e-learning purposes, leading to low loss ratios and high reliability. H.263 is especially appropriate for low bandwidth environments.

Table 1. vic and rat QoS parameters used to adjust applications’ profile



(1)

Rat

-f format

Indicates audio encoding format: l16, pcm, dvi, gsm and lpc

Vic

-B kbps

Sets the maximum bandwidth slider (kbps)

Vic

-c dither

On a color-mapped display, uses the algorithm indicated by dither (e.g., ed, gray, od, quantize) to convert to the available color palette

Vic

-f format

Indicates the video encoding format: h261, h263, jpeg, nv, ...

Vic

-F fps

Sets the maximum frame rate (fps)

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

E-lEarning sErVicEs anD FEaTUrEs The developed distance e-learning system presents numerous features providing distinct service levels, such as: 1.

2. 3. 4.

5. 6. 7.

Virtual academy, Web-based with refined usability, integrating authentication and services for the e-learning community; Registration, authentication and maintenance of educational agents; Multicast sessions maintenance and scheduling; Access to asynchronous material such as video on demand, slide presentations and other multimedia resources; Interactive multimedia multicast conferences with QoS adaptation; Other multicast tools for shared workspace; Discussion spaces such as forum and multicast chat room.

This information system incorporates online databases structuring courses, students, tutors and sessions’ data. These resources were developed using MySQL/PHP. A Web site congregating all developed application component prototypes is available at www.esa.ipb.pt/multicast. Certain processes for assessing hardware performance require user’s explicit authorization, allowing extended security privileges to applets in order to perform system’s inspection, collecting substantial data used by subsequent applets of the control path. The security certificates used in this work are not provided by official entities, but generated by applet compilation tools for testing. Although the adaptation process is totally transparent, effectively, the users may edit QoS parameters suggested by the system. If editing occurs, correctness and validation are assured by embedded Javascript code for parsing purposes. All MMC applications need to be previously installed and accessible through the command

Figure 3. E-learning system screenshots and MMCs adaptation HTML forms based on available QoS



Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

line interface, configuring the PATH environment variable properly. If we want to transmit audio or video the required equipment must also be ready. Gathering these basic requirements it is possible to participate in e-learning sessions, having adapted QoS in a transparent way, with great usability. Figure 3 integrates system’s screenshots, illustrating step by step, when a QoS adapted videoconference is selected from the “Services” menu.

PErFormancE sTUDY Applications that use voice, video streams, or multimedia must be carefully managed within an IP network to preserve their operational integrity. Beyond routing improvements, QoS in a multimedia conference needs primarily to deal with several sources with different characteristics, shifting large amount of traffic competing for network capacity. MMC applications may easily absorb all network resources and the subjective quality sensed by users would remain poor if the available resources are used indiscriminately. As mentioned earlier, the adaptation purpose, with e-learning in mind, was to integrate MMC applications with QoS conscience, preserving resources in order to maintain conference quality and improve scalability.

In order to test the framework, different scenarios were simulated and the corresponding resources’ consumption verified considering the QoS limitations associated with each QoS mode defined in Table 2. For videoconferencing, regarding e-learning purposes, it is widely accepted that reference values correspond to “Maximum quality, few action scene.” Bandwidth consumption in vic default mode is 128 kbps. When adaptation is requested, the different adaptation modes use the values charted in Figure 4. For instance, the best quality mode consumes around 400 kbps, allowing better image and motion. Different equipment was also tested in order to validate the rank of the defined adaptation modes. We observed that modern high performance equipment tends to be neutral, in this case adaptation will be influenced overall by network conditions, but with mobile computation in mind, PDAs and cellular phones, CPU performance should not be relegated. The experimental results were obtained varying the number of new multicast members transmitting voice and video, considering that all the multicast group members are multimedia receivers and transmitters capable using any source multicast (ASM) technology via vic and rat applications. Due to the limited number of multicast monitoring tools publicly available, we use embedded applications resource meters and

Table 2. Set of parameters for different QoS adaptation modes



ADAPTATION

MAXIMUM

FRAME RATE

VIDEO CODEC

COLOR

AUDIO CODEC

5

1 Mbps

30 fps

H.261

Yes

L16

4

512 kbps

25 fps

H.261

Yes

PCM

3

256 kbps

20 fps

H.261

Yes

DVI

2

128 kbps

15 fps

H.263

Yes

GSM

1

64 kbps

10 fps

H.263

No

LPC

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

Figure 4. BW needs for each QoS mode

Figure 5. CPU needs for each QoS mode

the Multicast Monitor (www.multicastmonitor. com) to collect and handle the resulting data. The e-learning system, more concretely the tested prototype, showed good performance indicators that validate the architecture model proposed. Because video traffic is quantitatively more representative of resource consumption, it was analyzed preferentially. Figures 4 and 5 exhibit the levels of resource consumption for each QoS mode considered. Here, the overhead introduced by middleware to prepare applications is marginal as it occurs before transmission time.

QoS adaptation, when treated systematically in tolerant real-time applications, denotes advantages in group scalability and QoS sustainability in heterogeneous and unpredictable environments such as the Internet and Mbone. Figures 6, 7 and 8 illustrate a comparison between two simulated sessions, the first without QoS adaptation and the second including adaptation managed by the developed middleware layer. The results show that scalability is increased, but equally important is the fact that applications may benefit from resource availability what does not occur when using the



Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

Figure 6. Linear bandwidth distribution using applications’ defaults, no adaptation is used

Figure 7. Increasing the number of active group members using adaptation to distribute network resources

Figure 8. QoS mode adopted by the system facing the available resource conditions

default applications’ configuration. When the available resources decrease, the system allocates them to critical parameters. For instance, while the frame rate should not be below 10 fps, the

0

image quality may be poor or monochromatic if the contents are correctly perceived. Limiting bandwidth to applications, not only with explicit parameterization but also choosing

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

the right encoding format for e-learning sessions, allows efficient resource utilization and proactive usability, avoiding network overload and congestion. If the network load remains high it is easier to recover if adaptation is used. The experience with Mbone showed that elearning groups tend to be small, usually less than twenty members. Effectively, e-learning communities, as in traditional training methods, need a tutor, who is mainly an educational agent and not necessarily a learning technologist. Indeed, questions related with communication technologies won’t constitute pedagogical limitation if intelligent QoS management autonomy is provided natively or by middleware to applications. E-learning conferencing specificity requires appropriate video encoding formats able to achieve low loss ratio and fast recovery from congestion. We compared H.261 and Motion JPEG (ITUJPEG, 1992) performance in experimental sessions, using a modest PC (PIII 0.7GHz - 256 MB RAM). The results were penalizing for MJPEG, where loss was about 30%, in opposition to 1% for H.261.

conclUsion anD DiscUssion The goal of this work was, in one topic, to foster “ecological” practices in the Internet when using MMCs applications in e-learning services. The proposed system integrates public domain multicast applications for synchronous media communication, being supervised by a middleware based QoS management framework, intending to preserve the QoS of critical parameters for e-learning session’s specificity. As main contributions, this work: 1.

2.

Provides an integrated e-learning environment based on interactive multimedia services with proactive QoS; Improves the usability of MMC applications;

3.

Allows the development of end-to-end QoSaware multimedia conferences, coordinating resources from network, end-system processing equipment and applications.

Middleware adaptation is a solution that suits the present state of Internet and the requirements of new multimedia distributed applications. We use a middleware layer to manage QoS adaptation in interactive audio and video applications coordinating resource demand, monitoring and adjudication. Substantive results were obtained in group scalability, QoS sustainability and proactive resource utilization. Comprising multiple sources (even unauthorized ones), ASM involves high complexity and may compromise the success of e-learning conferences. Future work includes the use of SSM (source-specific multicast) (Holbrook & Cain, 2006) in order to overcome this limitation. Multicast communications could also benefit from other implementation simplifications, namely the recently proposed FRM (free ride multicasting) (Ratnasamy, Ermolinskiy, & Shenker, 2006). The development of new multilayer video encoding formats could also increase the flexibility when using QoS adaptation. When cumulative layers are transmitted avoiding redundancy, using different SSM groups or channels, adaptation can be performed in a transparent way in order to achieve efficient resources utilization (Johanson & Lie, 2002; Liu, Li, & Zhang, 2004).

rEFErEncEs Allison, C., Ruddle, A., McKechan, D., & Michaelson, R. (2001). The architecture of a framework for building distributed learning environments. In Proceedings of the 2001 IEEE International Conference on Advanced Learning Technologies (pp. 29-32). Wisconsin: IEEE Press.



Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

Amoretti, M., Bertolazzi, R., Reggiani, M., Zanichelli, F., & Conte, G. (2006). Designing Grid services for multimedia streaming in an e-learning environment. Concurrency and Computation: Practice and Experience, 18(8), 911-923.. Bolot, J., Crépin, H., & Garcia, A. (1995). Analysis of audio packet loss in the Internet. In Proceedings of the 1995 Workshop on Network and Operating System Support for Audio and Video (pp. 163 -174). INRIA, France. Bouras, C., & Gkamas, A. (2003). Multimedia transmission with adaptive QoS based on real time protocols. International Journal of Communications Systems, 16, 225-248. Deering, S. (1998). Multicast routing in Internetworks and extended LANs. In Proceedings of ACM SIGCOMM (pp. 55-64). Deusdado, S. (2002). Integração Adaptativa de Aplicações Multicast para Conferência Multimédia. Master Thesis, Departamento de Informática, Universidade do Minho. Eklund, J., Kay, M., & Lynch, H. (2003). E-learning: Emerging issues and key trends. Australian Flexible Learning Framework, Australian National Training Authority. Hardman, V., Kirstein, P., Sasse, A., Handley, M., & Watson, A. (1995). RAT, Robust Audio Tool. Retrieved from http://www-mice.cs.ucl. ac.uk/multimedia/software/rat/ Holbrook, H., & Cain, B. (2006). Source-specific multicast for IP (RFC 4607, IETF). Network Working Group. ITU-JPEG. (1992). JPEG Standard, Information Technology - digital compression and coding of continuous-tone still images - requirements and guidelines. Recommendation T.81, ITU. ITU-T H.261. (1993). ITU-T Recommendation H.261: Video CODEC for audiovisual services At p x 64 kbits.



ITU-T H.263. (1998). ITU-T Recommendation H.263: Video coding for low bitrate communication. JMF 2.0 (1999). JMF - Java TM Media Framework API Guide. Sun Microsystems, JMF 2.0 FCS. Johanson, M., & Lie, A. (2002). Layered encoding and transmission of video in heterogeneous environments. ACM. Kosiur, D. (1998). IP multicasting: The complete guide to interactive corporate networks. New York: John Wiley & Sons, Inc. Layaida, O., & Hagimonte, D. (2002). Dynamic adaptation in distributed multimedia applications (Technical Report). INRIA. Li, B., Xu, D., Naharstedt, K., & Liu, J. (1998). Endto-end QoS support for adaptive applications over the Internet. Department of Computer Science, University of Illinois at Urbana-Champaign. Liu, J., Li, B., & Zhang, Y. (2004). An end-to-end adaptation protocol for layered video multicast using optimal rate allocation. IEEE Transactions On Multimedia, 6(1). Lubonski, M., Gay, V., & Simmonds, A. (2005). A conceptual architecture for adaptation in remote desktop systems driven by the user perception of multimedia. In Proceedings of the 2005 AsiaPacific Conference on Communications (pp. 891-895). McCanne S., & Jacobson, V. (1995). vic: A flexible framework for packet video. In Proceedings of the ACM Multimedia Conference (pp. 511-522). Retrieved from http://www-mice.cs.ucl.ac.uk/ multimedia/software/vic/ Miras, D. (2002). A survey of network QoS needs of advanced Internet applications (Working Document). Computer Science Department, University College London. Mohsin, M., Wong, W., & Bhutt, Y. (2001). Support for real-time traffic in the Internet, and

Synchronous E-Learning Integrating Multicast Applications and Adaptive QoS

QoS issues. Department of Computer Science, University of Texas at Dallas. Ratnasamy, S., Ermolinskiy, A., & Shenker, S. (2006). Revisiting IP multicast. In Proceedings of the ACM SIGCOMM 2006 Conference. Steeples, C., & Jones, C. (2002). Networked learning: Perspectives and issues. London, UK: Springer. Thaler, D., & Handley, M. (2000). On the aggregatability of multicast forwarding state. In INFOCOM 2000. Proceedings of the Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Vol. 3, pp. 16541663). IEEE. Tusch, R., Böszörményi, L., Goldschmidt, B., Hellwagner, H., & Schojer, P. (2004). Offensive and defensive adaptation in distributed multimedia systems. ComSIS, 1(1).

Vandalore, B., Feng, W., Jain, R., & Fahmy, S. (2001). A survey of application layer techniques for adaptive streaming of multimedia. Real Time Imaging, 7(3), 221-235. Wu, D., Hou, Y., Zhu, W., Zhang, Y., & Peha, J. (2001). Streaming video over the Internet: Approaches and directions. IEEE Trans. Circuits Syst. Video Technol., 11(3), 282-300.

EnDnoTE 1

RTT and/or P values will be, if necessary, assigned to 1 to avoid division by zero. To prevent incongruence, the maximum bandwidth allowed cannot exceed the detected value (B), otherwise the computed mode will suffer cyclic decrements while the excess remains and M>1.





Chapter II

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach: A Longitudinal Field Experiment Charlie C. Chen Appalachian State University, USA R. S. Shaw Tamkang University, Taiwan

absTracT The continued and increasing use of online training raises the question of whether the most effective training methods applied in live instruction will carry over to different online environments in the long run. Behavior modeling (BM) approach—teaching through demonstration—has been proven as the most effective approach in a face-to-face (F2F) environment. A quasi-experiment was conducted with 96 undergraduate students who were taking a Microsoft SQL Server 2000 course in a university in Taiwan. The BM approach was employed in three learning environments: F2F, online synchronous and online asynchronous classes. The results were compared to see which produced the best performance, as measured by knowledge near-transfer and knowledge far-transfer effectiveness. Overall satisfaction with training was also measured. The results of the experiment indicate that during a long duration of training no significant difference in learning outcomes could be detected across the three learning environments.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

concEPTUal FoUnDaTions The Internet’s proliferation creates a wealth of opportunities to deploy alternative online learning environments to facilitate many users in their learning processes. The information technology (IT) skills training market represented 76% of the entire online learning market in year 2000, according to a Jupiter Research report (CyberAtlas, 2003). The worldwide corporate online learning market may grow to $24 billion ($18 billion in the U.S.) by 2006 with a compound annual growth rate of 35.6% (IDC, 2002). The burgeoning online learning/training market, and the increasing training budgets of businesses and schools has provided these key users of online training and marketing tools with practical reasons, as well as compelling research motives, to investigate the effectiveness of training and education in different online formats. Online learning differs primarily from the traditional face-to-face (F2F) learning in that it is a user-centered, rather than instructor-centered, learning mode. Other benefits of substituting online learning for F2F learning include (1) selfpaced instruction; (2) the ability to incorporate text, graphics, audio and video into the training; (3) opportunity for high levels of interactivity; (4) a written record of discussions and instructions; (5) low-cost operation; and (6) access to a worldwide audiences (Aniebonam, 2000). In addition, online learning can remove a certain degree of space and time limitations, speed up the learning process for motivated learners, lower economic costs of attending F2F classes and have higher information accessibility and availability. Although IT has changed the training and educational approaches and environments, the ultimate goal of learning has not changed, that is, to transfer knowledge to students and allow them to apply the acquired knowledge in real situations. In the field of IT, the success of software training can be assessed with a trainee’s IT skills of, and knowledge of the use of, particular software

to solve problems. Surprisingly, after attending a training session, very few students know how to properly apply the acquired knowledge and skills to real situations. This raises an important issue, that is, how to improve knowledge transfer capability of learners in different online learning environments. The importance of knowledge transfer is selfevident. However, the knowledge transfer process does not occur naturally. There is a need to assist learners in transferring their acquired knowledge into future applications. One effective approach to assisting the learning transfer process is “behavior modeling” (BM). This approach teaches learners through demonstration and hands-on experience. Simon, Grover, Teng, and Whitcomb (1996) and Compeau and Higgins (1995) found that in the field of information technology, BM is the most effective approach compared to the other two knowledge transfer approaches: exploration—teaching through practice on relevant example, and instruction—teaching software characteristics. Distance education is defined as “teaching through the use of telecommunications technologies to transmit and receive various materials through voice, video and data” (Bielefield & Cheeseman, 1997, p. 141). In the same token, Leidner and Jarvenpaa (1995) define distance learning as “the transmission of a course from one location to another” (p. 274). These definitions provide an analogy to distance learning in the field of information technology or online software training. Online software training can be the transmission of instructional IT programming or contents to geographically dispersed individuals or groups. There are two general modes of online learning: synchronous and asynchronous modes. Each mode can be marshaled with IT tools to deliver software training. Case in point, audio and video conferences are two types of online synchronous training mediums. Online asynchronous training mediums range from Web pages, file download, e-



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

mail, e-mail list, newsgroup, forum, chat, response pad, whiteboard and to screen sharing. Built on his personal distance training and education experiences since 197 -+ 1, Horton (2000) suggests that online synchronous and asynchronous learning and training be designed for different purposes. Incorporating synchronous learning demands the control of schedule, time, people, class size, video and audio equipment and place. These factors constrain the possibility of reaching large numbers of students at any given time and in any given place. However, the BM approach for trainees can be a problem in online asynchronous and synchronous training. For instance, any demonstration presented by a live instructor would need to be replaced with a scripted or videotaped demonstration in asynchronous mode, and live transmission or Webcam in synchronous mode. This raises several important questions. Can the scripted, videotaped, live transmission or Webcam approach still be as effective as the traditional classroom? How receptive are students to different online learning environments with differential degrees of student-centered interaction compared to an instructor-centered F2F environment? Most importantly, it is an unknown but interesting question to ask whether knowledge can be effectively transferred in different online environments. This research is to address these important issues faced by any instructor who intends to apply the BM approach in either online synchronous and asynchronous environments.

bEhaVioral moDEling anD KnoWlEDgE TransFEr Social learning theory is the basis of the behavior modeling approach. Therefore, it is important to assess the applicability of the theory and approach in the online learning environment. Learning outcomes can be measured by different types of knowledge transfer and end-user satisfaction.



behavior modeling in online Environments Bandura (1977) proposed the Social Learning Theory to explain the interactive learning process between individuals and their social environment. He asserted a series of social learning needs take place to direct an individual from biological and self-centered response to social and group behaviors. Since the social learning process takes place within a society, individuals learn to establish their behavior models by observing and imitating other individuals’ behaviors or through the enforcement of the media and environment. Online learning in different environments needs to be delivered via different media. Different online learning environments, therefore, may have different degrees of enforcement to learners’ individual behaviors. Learning by modeling or observing people’s behaviors may be more effective than learning by trial-and-error because the former approach can avoid unnecessary mistakes and harm. Modeling an instructor’s behaviors empowers students to (1) learn new behavior from the instructor, (2) selfevaluate their behaviors against the instructor’s and (3) enforce students’ current behavior. Learning by modeling takes place in four sequential steps: (1) attention, (2) retention, (3) motor reproduction and (4) motivation and reinforcement (Bandura, 1977). Enforcement forces, such as the duration of training, praise, motivation and attention of others, allows learning to move along these four steps against counter forces. Enforcement forces, such as retention enhancement and practice, can contribute to better cognitive learning (Yi & Davis, 2001). Lewin (1951) argued that the effectiveness of Behavior Modeling is a function of people interacting within an environment. The BM approach is different from learning by adaptation. The former approach teaches through demonstration, while the latter approach influences the behaviors of learners by reward and punishment (Skinner,

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

1938). The BM approach was first applied in the training of interpersonal communication and management skills (Decker & Nathan, 1985). Gist, Schwoerer and Rosen (1989) further applied the training to the context of information technology. BM may be readily employed in face-to-face instruction, but cannot be easily simulated in online asynchronous instruction, which lacks the interactive immediacy necessary for optimally effective instructor demonstration and correction. The richness of information media in online synchronous instruction is another constraint and may also have less enforcement force than F2F instruction to the learning outcomes. For example, in a live training class, the instructor is able to demonstrate a software process and immediately ask the students to repeat the activity under the instructor’s close supervision. However, in an online asynchronous situation where there is no live instructor, the demonstration loses the benefit of that immediate feedback. In the same token, in an online synchronous situation bandwidth constraints and compromised reciprocity may undermine the enforcement force of the demonstration. In both online environments, enforcement forces can be further compromised with the missing of “learning by doing,” another key element of F2F BM training (McGehee & Tullar, 1978). Therefore, there is a strong possibility that the BM approach cannot be fully replicated in either the online synchronous or asynchronous situation and will not be as effective a method in online training as in the traditional environment.

Knowledge Transfer Knowledge transfer is the application of acquired skills and knowledge into different situations. Unless the transferring process occurs, learning has little value. The applied situations could be similar or novel to the learning situation. Depending on the situation, knowledge transfer can take place

in different formats. In general, there are four different types of knowledge transfer.

Positive Transfer vs. Negative Transfer Positive transfer of learning means that learning in one situation stimulates and helps learning in another situation. Negative transfer of learning hinders the application of learning in one situation to other situations. Positive learning experience can be enhanced via analogy, informed instruction (Paris, Cross & Lipson, 1984), tutorial (Morris, Shaw & Perney, 1990) and so forth. Learning effectiveness can be improved by triggering positive learning and mitigating negative learning experience.

Near Transfer vs. Far Transfer Salomon and Perkins (1988) argued that transfer of learning could have a differential degree of transfer. The effectiveness of near-transfer learning depends on the learner’s ability to solve problems similar to those encountered in the learning context. For instance, learning how to add two digit numbers allows learners to add three digit numbers. Near-transfer learning occurs in two similar situations and at a lower level. Therefore, the level of learning is more easily acquired and applied. In contrast, applying the acquired skills and knowledge in two dissimilar and sometimes novel situations is much harder to achieve. For instance, a table tennis player can apply skills of playing pinball to playing tennis. Although both sports look similar on the surface, the techniques to control pinballs and tennis balls are very different. The learning transfer is much harder to be acquired and retained. Therefore, the transfer is defined as far-transfer learning. Near-transfer and far-transfer of knowledge seem to be the most widely used measures of learning outcomes in the field of information technology since learners must utilize the knowledge learned in a computing environment.



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Specific Transfer and General Transfer Depending on learning content, there are two different learning transfers: specific transfer and general transfer (Bruner, 1996). The former refers to the extension and association of habit and skills. The latter refers to the transfer of principles and attitudes that can be used to deepen the understanding of basic concepts.

Lateral Transfer and Vertical Transfer Gagne (1992) asserted that the transfer of learning includes lateral and vertical transfers. Lateral learning is to apply one domain of knowledge to another domain. Lateral learning does not follow step-by-step instruction and is considered as provocative learning. Vertical learning means that a higher level of learning needs to be created by integrating acquired skills, and experiences with new situations. Vertical transfer of learning is analytical and sequential.

Other Knowledge Transfer Theories Theories related to knowledge transfer are not limited to the above mentioned ones. For instance, the theory of identical elements asserts that the more identical elements different learning contains, the more efficient the transfer of learning (Thorndike, 1949). Baldwin and Ford (1988) proposed a general training theory to classify three categories of factors affecting transfer of training: (1) training inputs, (2) training outputs and (3) conditions of transfer. The situated learning theory argues that individuals are affected by learning environment when trying to solve practical problems. Therefore, the interaction between learners and the environment is an important factor that needs to be taken into account when measuring the transfer of learning. Finally, the theory of formal discipline argues that knowledge transfer skills can be acquired by training learner’s sensuality, such as



thinking, judgment, classification, imagination, creation and so forth. The objective of this study was to investigate the impacts of the learning environment in online and offline formats on the transfer of learning. The situational changes rationalize the adoption of situated learning theory. To accomplish this objective, we sought to train end-user to learn how to use Microsoft SQL server 2000 software. Therefore, we adopted the near-transfer and fartransfer measures of learning outcomes for our information technology related experiment.

hYPoThEsEs Hypotheses are formulated to investigate whether the BM approach is as effective in online synchronous and asynchronous environments as in the traditional face-to-face environment. We measured learning outcomes by trainees’ performances in near-transfer and far-transfer tasks, as well as overall satisfaction levels. The study also considered the importance of time variant. Hence, training and performance measurement were conducted over five weeks.

Knowledge near-Transfer (KnT) Tasks H1: End-users trained using F2F behavior modeling perform near-transfer information system tasks better than those trained in asynchronous behavior modeling. H2: End-users trained in F2F behavior modeling perform near-transfer information system tasks better than those trained in synchronous behavior modeling. H3: End-users trained in synchronous behavior modeling perform near-transfer information system tasks better than those trained in asynchronous behavior modeling.

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Knowledge Far-Transfer (KFT) Tasks

subjects control

H4: End-users trained in F2F behavior modeling perform far-transfer information system tasks better than those trained in asynchronous behavior modeling. H5: End-users trained in F2F behavior modeling perform far-transfer information system tasks better than those trained in synchronous behavior modeling. H6: End-users trained in synchronous behavior modeling perform far-transfer information system tasks better than those trained in asynchronous behavior modeling.

The setting for the field experiment was the Tamkang University in Taiwan. The experiment was prompted by the need of 96 college sophomores, who are Management Information Systems (MIS) majors, to learn a Microsoft SQL Server 2000 software program in a database processing course. The schedule agreed on with the faculty at Tamkang University was to run the experiment for an hour training each week for four weeks. The author’s graduate assistant Ms. Lin helped administer the experiment to collect the data. The subject pool had a mean age of 22 years. Subjects who participated in the structured experiment had little database-related experience. Their intellectual levels are relatively the same because subjects scored the same range of scores in a national entrance exam. The national entrance exam system has been adopted for more than 40 years in Taiwan and is considered a relatively reliable test. Subjects’ individual backgrounds should not have influence on learning outcomes. For the purposes of this study, subjects were chosen if they lacked a theoretical and procedural understanding of the particular subject area being tested. Participants were given a pretraining questionnaire that includes important study units on Microsoft SQL Server 2000. Two experts of the domain administered the Delphi study to finalize the study units and questionnaires. This is to improve the content validity. The subjects voluntarily answered whether they knew those study units and answered their database-related experiences. Based on their answers, a correlation test of database and usage experience of the target system showed no significant differences among three experimental groups. Subjects of the study may be considered representative of novice end-users. Many studies (Ahrens & Sankar 1993; Santhanam & Sein, 1994) support using students as experimental subjects to represent the general populations. Hence, all subjects’ questionnaires were used for further data analysis. This segmenta-

overall satisfaction H7: End-users trained in synchronous behavior modeling have a higher overall satisfaction level than those trained in asynchronous behavior modeling.

rEsEarch DEsign This study applied Simon, Grover, Teng and Whitcomb’s (1996) well-constructed software training theory to experimentally test behavior modeling training in three learning environments—F2F, online asynchronous and online synchronous environments. In doing so, it should be possible to detect the effects of the single independent variable (training environment) on training outcomes. The experiment was conducted in a field setting that enabled the study to garner greater external validity than would be the case with a laboratory experiment. A field experiment methodology has the merits of “testing theory” and “obtaining answers to practical questions” (Kerlinger & Lee, 2000). The exploratory nature of the study requires that variables (e.g., training environments and subject areas of study) under investigation be manipulated.



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

tion was used to mitigate the effects of computer literacy and experience on the findings, thereby improving the internal validity of the study.

designed to maximize the effect of size on their differences (Figure 1).

Training Procedures Training Treatments Face-to-face BM (FBM) is instructor-centered training while online asynchronous BM (ABM) and synchronous BM (SBM) are learner-centered training. Course materials used in online learning environments were created to properly reflect the key elements of a behavior modeling approach. AniCam simulation software was used to record the demonstration of instruction. Hyperlink structure was used to help users assimilate nonlateral conceptual, and procedural knowledge. Feedback activities of behavior modeling approach in online asynchronous environment are supported with e-mail and hyperlinks. SBM differs from ABM in providing feedback functions via real-time discussion forums. Training materials integrate key elements of behavior modeling approach: (1) control of three different learning environments, (2) demonstration of the instructor, (3) continuous feedback (verbal feedback in F2F and online synchronous environments; e-mail feedback in the online asynchronous environment). Three training environments were

The experimental study lasted for four weeks. There was a 50 minute training session each week for each class. Figure 2 shows the experimental procedures used at each time period. The X’s, Y’s and Z’s represent online asynchronous BM training, online synchronous BM training and F2F BM training methods, respectively. The subscripts next to each alphabet indicate the ith observation or training session, respectively. Before executing experimental treatments (the pretest period O1), the instructor asked the subjects to complete a short questionnaire soliciting demographic information, database software-related experience and attitudes towards learning in the subject’s assigned online learning environment (Pretest). Approximately one-third of the subjects pooled received the same experimental treatment for four straight weeks (Week1 to Week4). The assigning process was random on the class basis. Randomizing the execution of O4 and O5 in Week2 and Week3 for Group A and Group B can help avoid possible confounding results from the interactive effects of the pretest of O1 and O3. This random-

Figure 1. Differences of behavior modeling approach in three learning modes Online Learning Environments Asynchronous BM (ABM) • Scripted demonstration of step-by-step instructions • Deductive/inductive complementary learning • Trainees choose one of two relevant examples to practice • Without online reference sources • Trainee control

0

Synchronous BM (SBM) • Webcam-delivered demonstration of step-bystep instructions • Deductive/inductive complementary learning • Instructor chooses examples that are relevant to trainees’ majors • Without online reference sources • Trainer/trainee partially control

Off-line Learning Environment Face-to-Face BM (FBM) • Demonstration of a live instructor to learn step-by-step • Deductive/inductive complementary learning • Live instructor chooses examples that are relevant to trainees’ majors • Without online reference sources • Trainer control

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Figure 2. Experimental procedures GROUP

Pretest

Week1

Week2

Week3

Week4

Post-test

Group A

O1

O2 X1 O3

X2

O4 X3 O5

X4

O6

Group B

O1

O2 Y1 O3

O4 Y2 O5

Y3

Y4

O6

Group C

O1

O2 Z1 O3

O4 Z2 O5

Z3

Z4

O6

Oi ＝ Xi ＝ Yi ＝ Zi ＝

Questionnaire and Tests ABM (Online Asynchronous BM Training) SBM (Online Synchronous BM Training) FBM (F2F BM Training)

Figure 3. Delivery mechanisms of behavior modeling approaches FBM (F2F Behavior Modeling)

ABM (Asynchronous Behavior Modeling)

SBM (Synchronous Behavior Modeling)

Instructor demonstrates the use of software along with PowerPoint slides

Course materials covered by FBM was pre-recorded and stored in a server.

Instructor was present, but broadcasted steaming video from a broadcast room.

Covered three study subjects within forty five minutes each week

No instructor was present to assist the learning process of students. Students learned at their own path and completed their study within forty five minutes.

Instructor conducted the realtime discussion with students on a BBS station.

Information Systems Tools

Instructor, PowerPoint, and PC

AniCam, PowerPoint and Acrobat Reader

AniCam, PowerPoint, Stream Author v.2.5 (Authoring Tool) and Acrobat Reader

Target System

SQL Server2000 Personal Edition

SQL Server2000 Personal Edition

SQL Server2000 Personal Edition

Pretest Questionnaire

Learning Experience and Style Questionnaires

Learning Experience and Style Questionnaires

Learning Experience and Style Questionnaires

The First Week

First Training Session First Learning Outcomes Test

First Training Session First Learning Outcomes Test

First Training Session First Learning Outcomes Test

The Second Week

Second Training Session Second Learning Outcomes Test

Second Training Session

Second Training Session Second Learning Outcomes Test

The Third Week

Third Training Session

Third Training Session Second Learning Outcomes Test

Third Training Session

The Fourth Week

Comprehensive Test (Third Learning Outcomes Test)

Comprehensive Test (Third Learning Outcomes Test)

Comprehensive Test (Third Learning Outcomes Test)

Post-test Questionnaire

Measure End-User Satisfaction

Measure End-User Satisfaction

Measure End-User Satisfaction

Course Materials

ization process can further ensure that difference in learning outcomes of O6 is not possibly due to the sensitization of the participants after the pretest and the interaction of their sensitization, O4 and O5 (Kerlinger & Lee, 2000).

Before or after each training session, subjects were asked to complete database design tasks using the MS SQL commands to assess their prior knowledge in the trained subjects and immediate learning outcomes that involve both near-transfer 

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

and far-transfer knowledge. On week five, students were evaluated again for their attitude changes towards the e-learning sessions and performance in near and far-transfer tasks (Post-test). The final exam concludes the five-week training sessions. Training materials were designed to integrate key elements of the three training environments, as illustrated in Figure 3. Course materials used in the online asynchronous training session were stored on the school’s server for students to learn at their own pace after each training session was completed. At the end of the experiment, students were asked about their affect for their learning environments.

outcomes measurement Regardless of the teaching environment, computer training is intended to instill in users a level of competency in using the system and to improve their satisfaction with the system. A user’s competency in using a system is contingent upon the user’s knowledge absorption capacity. Ramsden (1988) finds that effective teaching needs to align students with situations where they are encouraged to think deeper and more holistically. Kirkpatrick (1967) also suggests that learning effectiveness needs to be evaluated by students’ reactions, learning and knowledge transfer. The levels of knowledge absorbed by students, Bayman and Mayer (1988) suggest, may include syntactic, semantic, schematic and strategic knowledge. Mennecke, Crossland and Killingsworth (2000) believe that experts of one particular knowledge domain possess more strategic and semantic knowledge than novices. Knowledge levels, as Simon, Grover, Teng and Whitcomb (1996) suggest, can be categorized as near-transfer, far-transfer or problem solving. Near-transfer knowledge is necessary for being able to understand software commands and procedures. This type of knowledge is important for a trainee to be able to use software in a step-by-step fashion. Far-transfer



knowledge seeks to ensure that a trainee has the ability to combine two or more near-transfer tasks to solve more complicated problems. Both the use of software and information systems and the satisfaction levels of using them are useful surrogates to properly measure the effectiveness of an information system (Ives, Olson, & Baroudi, 1983). The end-user satisfaction level has been widely adopted as an important factor contributing to the success of end-user software training. Since the study was to replicate Simon, Grover, Teng and Whitcomb’s (1996) research in a dissimilar environment, near-knowledge and far-knowledge transfer, and end-user overall satisfaction levels were adopted in this study to measure training outcomes. Cronbach’s alpha reliability for Simon et al.’s (1996) instrument to measure satisfaction is r = 0.98. Users need to use the Likert scale from one to five to answer 12 test items related to their satisfaction with the use of online system.

DaTa analYsis Table 1 shows the means and standard deviations for the scores at each treatment period. Table 2 shows F and P values of the dependent variables (near-transfer and far-transfer task performances, and overall satisfaction) across treatment groups and in different times. Pretest scores (Q1, Q2 and Q4) in varying weeks were used to tell apart students with prior experiences and knowledge on the studied topics. After learning in a weekly session, students participated in a post-test. Their scores (Q3 and Q5) were used for KNT effectiveness comparison across training sessions. Scores of Q6 are KFT effectiveness and end-user satisfaction levels. A cursory examination of means (Table 1) indicates that no patterns can be identified for near-transfer performance from time Week1 to Week5. Subjects in ABM performed better than those in FBM, followed by SBM at Week1 while

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Table 1. Descriptive statistics: Means (standard deviations) ABM (N=40)

SBM (N=26)

FBM (N=30)

Overall (N=96)

KNT (Week 1)

27.63 (5.77)

25.38 (7.06)

26.00 (7.24)

26.51 (6.61)

KNT (Week 2)

28.50 (3.43)

28.46 (3.09)

29.83 (0.91)

28.91 (2.83)

KNT (Week 3)

56.05 (14.06)

64.27 (17.52)

62.27 (12.71)

60.22 (14.98)

KNT (Week 5)

71.70 (16.21)

70.50 (19.78)

67.00 (17.27)

69.91 (17.49)

KFT (Post-test)

9.63 (1.33)

9.42 (1.63)

8.83 (2.15)

9.32 (1.72)

OS (Post-test)

38.10 (6.87)

39.38 (9.21)

38.61 (7.83)

Table 2. Performance on differentlLearning outcomes over five weeks F

p-value

Power

KNT (Week 1)

1.035

0.359

0.337

KNT (Week 2)

2.415

0.095*

0.605

KNT (Week 3)

2.891

0.061*

0.677

KNT (Post-test)

0.634

0.532

0.246

KFT (Post-test)

1.913

0.153

0.517

OS (Post-test)

0.420

0.519

0.169

at Week2 and Week3 the order was changed to FBM>ABM>SBM and SBM>FBM>ABM, respectively. The findings are not in agreement with a consistent pattern as predicted by Hypotheses H1 and H2. For KFT tasks, subjects in ABM performed better than those in SBM, followed by FBM. This is the reversed order of a pattern as predicted by Hypotheses H3 and H4. The measurement of overall satisfaction level somewhat follows the predicted patterns of Hypotheses H5 and H6. We took a closer look at the mean difference at the significance level of 0.05. The study used one-way ANCOVA to analyze the effects of behavior modeling approach on learning outcomes over different time. Levene’s Test (1960) was used to examine the variance homogeneity of three groups. Its F-statistics showed that KNT was 3.04 (p=0.053) at Week1, 13.01 (p=0.000) at Week2, 1.71 (p=0.187) at Week3, and 0.47 (p=0.627) at the Post-test. In contrast, the F-sta-

tistics of Levene’s Test for KFT and OS were 7.64 (p=0.001) and 1.75 (p=0.191) at the Post-test. With the exception of KNT at Week2 and KNF at the Post-test, all dependent variables met the p > 0.05 criterion for assuring homogeneity of variances. The heteroscedasticity of variances for these two exceptions suggested that the statistical test results may not be valid. As such, the following discussion will ignore these two variances and focus on KNT and OS. For other effects that show significance, the study adopts the Scheffe post-test to analyze data. In addition, Pearson Correlation Analysis was used to assess the carry-over effects of different training sessions. ANCOVA was performed using the general linear model approach; the results are presented in Table 3. It shows that the treatment effects are significant for KNT (Week2) and KNT (Week3) with F-statistics of 2.415 (p=0.095) and 2.891 (p=0.061), confirming a univariate treatment effect of learning environments on the dependent



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Table 3. Results for training methods Variable

Hypothesis

Result in Correct Direction?

Week1

Week2

Week3

Week4

Post Test

Significant p-value? (n.s. - not significant)

H1: FBM > ABM

F

n.s. (p=0.312)

H2: FBM > SBM

T

n.s. (p=0.729)

H3: SBM > ABM

F

n.s. (p=0.182)

H1: FBM > ABM

T

p=0.051

H2: FBM > SBM

T

p = 0.069

H3: SBM > ABM

T

n.s. (p=0.956)

H1: FBM > ABM

T

p=0.083

H2: FBM > SBM

F

n.s. (p=0.612)

H3: SBM > ABM

T

p = 0.029

H1: FBM > ABM

F

n.s. (p=2.71)

H2: FBM > SBM

F

n.s. (p=0.459)

H3: SBM > ABM

F

n.s. (p=0.787)

H4: FBM > ABM

F

p=0.057

H5: FBM > SBM

F

n.s. (p=0.2)

H6: SBM > ABM

F

n.s. (p=0.639)

H7: SBM > ABM

F

n.s. (p=0.579)

(KFT)

Post Test (OS)

variable: KNT. However, the treatment effects are not salient for other dependent variables: KFT and OS. These lacks of effect may have been due to small effect sizes. Least-squares deconvolution (LSD) was used to test cross-correlations for KNT (Week2) and KNT (Week3). LSD is a cross-correlation technique for computing average profiles. LSD is very similar to most other cross-correlation techniques, though slightly more sophisticated in the sense that it cleans the crosscorrelation profile from the autocorrelation profile (Donati, 2003). For KNT (Week



2), the LSD results indicate that subjects in FBM perform better than those in ABM (p=0.051) and SBM (p=0.069). This supported the Hypotheses 1 and 2. However, Hypothesis 3 cannot be supported because the mean difference between ABM and SBM is not significant. For KNT (Week3), the LSD results indicate that (1) subjects in FBM performed better than those in ABM (p=0.083), and (2) subjects in SBM performed better than those in ABM (p=0.029). Hypotheses 4 and 6 are supported. Worthy to be noted is that H4 is upheld but in the reversed direction. This indicates that

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

ABM is a more effective method than FBM at improving knowledge far transfer. Four out of nine hypotheses in total are supported. Although not all hypothesized relationships are fully supported, the results obtained are interesting. The most intriguing result is that although there is statistically-justified reason for preferring FBM to ABM or SBM or software training, the pattern of results is not persistent in the long run. FBM resulted in better outcomes than ABM and SBM at Week2, and than ABM at Week3 for KNT. Although it never does so at a statistically significant level, subjects in ABM performed better than those in SBM, followed by those in FBM for KNT (Post-test) and KFT (Post-test). One interpretation of this is that either ABM or SBM training is no worse than FBM training across all dependent variables. The pattern of results for FBM suggests that trainers might choose ABM or SBM, which should to be a less costly alternative to FBM, without making any significant sacrifices in either learning or trainee reaction outcomes. Another result of interest is that, with respect to the three online asynchronous training methods, the pattern of results suggests that FBM might be the best for KNT in the short term. Of the nine hypotheses concerning relationships between these methods, four are in the expected direction, and significantly so. This indicates that use of ABM or SBM may be a better—and certainly no worse–software training strategy in the long term.

imPlicaTions For rEsEarch This chapter studied the impact of training duration on performance and trainee reactions. Trainees were exposed to the same training methods with different degrees of social presence for different durations. These findings indicated that training

duration and social presence have little impacts on learning outcomes. Despite this, the findings here raise additional questions for research. It may be more important to investigate the impacts of information richness (Fulk, 1993) features of online training media on training outcomes. Future studies might vary the social presence features of training media or their combination with social presence features (e.g., with instructor’s feedbacks versus discussion boards, e-mail response or playback features). Information richness may be a more influential factor affecting the performance of training approaches. It may also be useful to replicate the experimental equivalence of FBM, ABM and SBM methods of software training with different software and subjects. Since in the long term different treatments have similar impacts on learning outcomes, it may be practical to demonstrate the cost-based advantage of ABM over SBM, and SBM over FBM for software training in practical settings. Another way to improve the reliability of the study is to manipulate some useful blocking variables. A series of comparative studies can be conducted to assess the impact of individualism as a cultural characteristic, computer self-efficacy, task complexity (simple tasks vs. fuzzy tasks), professional backgrounds and the ratio of the training duration to the quantity of information to be processed, among others. Learning style may be an important factor to consider in the online learning environment. According to social learning theory, learners interact with the learning environment to change their behavior. Learning style is situational and can vary with different learning environments. Therefore, it is possible that the combination of training methods, learning style and social presence information richness (SPIR) attributes may jointly determine learning outcomes. This is not the case for BM approach in F2F environment. The self-paced online learning environment may alter the assertion. Hence, it may be necessary



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

to conduct longitudinal studies of the influence of learning style on learning performance and trainee reaction.

imPlicaTions For PracTicE The largest implication for practice is that ABM and SBM may provide cost-effective substitutes for FBM without significant reductions in training outcomes in the long term. While it may still be true that FBM is still the most effective approach to improve KNT in the short term, ABM and SBM have similar leverage in KFT in the short term and KNT in the long term. Regardless of training environments, trainees have same satisfaction levels in the near- and long-term. These findings strongly indicate that the cost issue is more important than learning effectiveness. When given the options to decide which BM approach to take in the long term, nonperformance issues (teacher and facility availability, trainee’s preferences, location and convenience issues) have to be first taken into account.

conclUsion The success of an online training strategy depends on its effectiveness in improving learning outcomes. This study, built on well-accepted frameworks for training research (Bostrom, Olfman & Sein, 1990; Simon & Werner 1996), examines the relative effectiveness of the behavior modeling approach in online synchronous, online asynchronous and face-to-face environments. The results from this experiment provide an empirical basis for the development of an online behavior modeling strategy: (1) FBM is more effective than ABM and SBM for knowledge transfer in the short term (KNT), and (2) ABM and SBM are as effective as FBM for knowledge transfer and overall satisfaction in the long term (KFT).



What is learned from this study can be summarized as follows: When conducting software training, it may be almost as effective to use online training (synchronous or asynchronous) as it is to use a more costly face-to-face training in the long term. In the short term face-to-face knowledge transfer model still seems to be the most effective approach to improve knowledge transfer in the short term. The limitation of this experimental study is that it was conducted with a homogeneous group with Taiwanese cultural and educational backgrounds. Therefore, this study may be constrained with the generalizability of its findings to different cultural contexts. Hofstede (1997) stated that the domains of education, management and organization have nurtured the values context that differs from one country to another. Cultural influences have been discerned in the study of Internet usage (Lederer, Maupin, Sena & Zhuang, 2000; Moon & Kim, 2001; Straub, 1997) and Web site design (Chu, 1999; Svastisinha, 1999). Users from different cultures have different perceptions about the usefulness and ease of use regarding different information systems (Straub, 1994). E-learning systems may differ based on the cultural backgrounds of the learners to improve their satisfaction levels and cognitive gains. Benefits of the congruence may include the improvement of (1) global e-learning adoption rate and (2) learning outcomes (attitude and cognitive gains). From the perspective of research design (Kerlinger & Lee, 2000), a cross-cultural study to replicate the study with American or European subjects may further validate and extend the generalizability of the findings. The study has accomplished its major goal; it provides evidence as to the relative effectiveness of the behavior modeling approach in different learning environments for software training. This research somewhat improves the generalizability of theories on the behavior modeling approach in different learning environments.

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

rEFErEncEs Ahrens, J. D., & Sankar, C. S. (1993). Tailoring database training for end users. MIS Quarterly, 17(4), 419-439. Aniebonam, M. C. (2000, October). Effective distance learning methods as a curriculum delivery tool in diverse university environments: The case of traditional vs. historically black colleges and universities. Communications of the Association for Information Systems, 4(8), 1-35. Baldwin, T.T., & Ford, J.K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41, 63-105. Bandura, A. (1977). Social learning theory. Morristown, NJ: General Learning Press. Bayman, P., & Mayer, R. E. (1988). Using conceptual models to teach BASIC computer programming. Journal of Educational Psychology, 80(3), 291-298. Bielefield, A., & Cheeseman, L. (1997). Technology and copyright law. New York: Neal-Schuman Publishers, Inc. Bostrom, R. P., Olfman, L., & Sein, M. K. (1990). The importance of learning style in end-user training. MIS Quarterly, 14(1), 101-109. Bruner, J. (1996). Toward a theory of instruction. New York: Norton. Chu, G.-L. (1999). The relationships between cultural differences among American and Chinese university students and the design of personal pages on the World Wide Web. Unpublished doctoral dissertation, University of Georgia. Compeau, D. R., & Higgins, C. A. (1995). Application of social cognitive theory to training for computer skills. Information Systems Research, 6(2), 118-143.

CyberAtlas. (2003). E-Learning market expanding beyond IT training. Jupiter Research. Retrieve July 28, 2006, from http://cyberatlas.internet.com/ markets/education/article/0,,5951_914901,00. html Decker, P. J., & Nathan, B. R. (1985). Behavior modeling training. New York: Praeger. Donati, J. (2003). Least-squares deconvolution of Stellar Spectra. Retrieved July 28, 2006, from http://webast.ast.obs-mip.fr/people/donati/multi. html Fulk, J. (1993). Social construction of communication technology. Academy of Management Journal, 36, 921-950. Gagne, R. M. (1992). Principles of instructional design. New York: Holt, Rinehart and Winston, Inc. Gist, M. E., Schwoerer, C., & Rosen, B. (1989). Effects of alternative training methods on selfefficacy and performance in computer software training. Journal of Applied Psychology, 74, 884-891. Hofstede, G. (1997) Cultures and organizations: Software of the mind. New York: McGraw-Hill. Horton, W. (2000). Designing Web-based training: How to teach anyone anything anywhere anytime. New York: John Wiley & Sons. IDC. (2002, September 30). While corporate training markets will not live up to earlier forecasts, IDC suggests reasons for optimism, particularly e-learning. Retrieved July 28, 2006, from http:// www.idc.com/getdoc.jhtml?containerId=pr2002_ 09_17_150550 Ives, B., Olson, M., & Baroudi, S. (1983). The measurement of user information satisfaction. Communications of the ACM, 26, 785-793. Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral research. New York: Harcourt Brace College Publishers.



Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Kirpatrick, D. L. (Ed.). (1967). Evaluation of training: Training and development handbook. New York: McGraw-Hill. Lederer, A. L., Maupin, D. J., Maupin, M. P., Sena, M.P. & Zhuang, Y. (2000). The technology acceptance model and the World Wide Web. Decision Support Systems, 29, 269-282. Leidner, D. E., & Jarvenpaa, S. L. (1995). The use of information technology to enhance management school education: A theoretical view. MIS Quarterly, 19, 265-291. Levene, H. (1960). In I. Olkin et al. (Eds.) Contributions to probability and statistics: Essays in honor of Harold Hotelling. (pp. 278-292). Stanford University Press.

prehension. Journal of Educational Psychology, 7, 1239-1252. Ramsden, P. (Ed.). (1988). Context and strategy: Situational influences on learning. In Learning strategies and learning styles. New York: Plenum Press. Salomon, G., & Perkins, D. N. (1988). Teaching for transfer. Educational Leadership, 46(1), 22-35. Santhanam, R., & Sein, M. K. (1994). Improving end-user proficiency: Effects of conceptual training and nature of interaction. Information Systems Research, 5(4), 378-399.

Lewin, K. (1951). Field theory in social science: Selected theoretical papers. New York: Harper and Row.

Simon, S. J., Grover, V., Teng, J. T. C., & Whitcomb, K. (1996). The relationship of information system training methods and cognitive ability to end-user satisfaction, comprehension, and skill transfer: A longitudinal field study. Information Systems Research, 7(4), 466-490.

McGehee, W., & Tullar, W. (1978). A note on evaluating behavior modification and behavior modeling as industrial training techniques. Personal Psychology, 31, 477-484.

Simon, S. J., & Werner, J. M. (1996). Computer training through behavior modeling, self-paced, and instructional approaches: A field experiment. Journal of Applied Psychology, 81(6), 648-659.

Mennecke, B. E., Crossland, M. D., & Killingsworth, B. L. (2000). Is a map more than a picture? The role of SDSS technology, subject characteristics, and problem complexity on map reading and problem solving. MIS Quarterly, 24(4), 601-627.

Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. New York: AppletonCentury Company, Incorporated.

Moon, J., & Kim, Y. (2001). Extending the TAM for a World Wide Web context. Information & Management, 38, 217-230. Morris, D., Shaw, B., & Perney, J. (1990). Helping low readers in grades 2 and 3: An after-school volunteer tutoring program. The Elementary School Journal, 91, 133-150. Paris, S. G., Cross, D. R., & Lipson, M. Y. (1984). Informed strategies for learning: A program to improve children’s reading awareness and com-



Straub, D. W. (1994). The effect of culture on IT diffusion: E-mail and FAX in Japan and the U.S. Information Systems Research, 5(1), 23-47. Straub, D., Keil, M., & Brenner, W. (1997). Testing the technology acceptance model across cultures: A three country study. Information and Management, 33, 1-11. Svastisinha, R. W. (1999). Wahhn: Web-based design. Wind and human comfort for Thailand. Unpublished doctoral dissertation, University of Southern California. Thorndike, R. L. (1949). Personnel selection: Test and measurement techniques. New York: John Wiley & Sons.

Online Synchronous vs. Asynchronous Software Training Through the Behavioral Modeling Approach

Yi, M. Y., & Davis, F. D. (2001). Improving computer training effectiveness for decision technologies: Behavior modeling and retention enhancement. Decision Sciences, 32(3), 521-544.

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 4, edited by S.-K. Chang and T. K. Shih, pp. 88-102, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).



0

Chapter III

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment Charlie C. Chen Appalachian State University, USA Albert L. Harris Appalachian State University, USA Rong-An Shang Soochow University, Taiwan

absTracT Case method teaching is prominent in its efficacy at improving the cognitive learning process in faceto-face classes. However, the efficacy of conducting this teaching method in an online asynchronous environment, where learners and instructor do not have real-time interactions, could be problematic. This study assesses and compares the efficacy of case method teaching in face-to-face and online asynchronous learning (OAL) environments. We proposed four hypotheses on the correlation between these two delivery modes of case method teaching and the learning performance of students. This study reports additional findings on the usage behavior of students in an online asynchronous environment. These findings are a useful aggregated surrogate to measure the effect of case teaching method in the online asynchronous environment. The overall findings of this study indicate that an online asynchronous environment can promote students’ participation in certain cases. As most antagonists for the adoption of online asynchronous case method surmised, cognitive learning gains via this learning method do not seem to be as high as in the face-to-face environment. The findings provide ample room for a further exploration of creative online asynchronous methods to continuously improve cognitive gains of learners. Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

inTroDUcTion The growing acceptance of e-learning technologies in the business school has promoted the rapid growth of online business degree offerings. Ninety-six online business degrees are available as of January 10, 2007, based on a cursory check of U.S. News E-learning Guide (2007). Some of these programs use a pure online format, while the others use a click-and-mortar format. The online asynchronous delivery of classes is location-free and time-free. As such, this delivery mode is becoming the most attractive learning method to online and part-time business students as opposed to the other learning modes (e.g., F2F and online synchronous modes). A growing number of students are turning to the online learning environment because of its convenience and accessibility. Case-based teaching is one of most effective methods to help business students acquire business concepts and address complex business problems. However, many antagonists have cast doubts on the efficacy of delivering case method teaching via online asynchronous learning environment. A wide variation of online asynchronous initiatives to improve the efficacy of case teaching method is being adopted. For instance, many universities in Taiwan are incorporating an online asynchronous gaming & simulation environment to improve the quality of their financial engineering programs (Wang, 2006). Active involvement is one important element that contributes to the success of case teaching method. To increase active involvement, the University of Exeter utilizes an online asynchronous learning system to enable leaders of student groups to create logbooks to facilitate the knowledge exchange and sharing process in a leadership class (Witzel, 2005). In recognition of the usefulness of delivering case teaching method via online learning technologies, Harvard Business School began enforcing the policy of having its students take a six-hour tutorial to acquire IT skills and concepts as a

prerequisite requirement to take core case-based courses (Bradshaw, 2006). Integrating online asynchronous learning technology into case teaching method poses many challenges to professors and students. First, online asynchronous learning systems can reach a global audience at anytime and from anywhere. Does this mean that class size is no longer an issue when conducting an online asynchronous case teaching class? It is still prevalent to see a limitation of class size in the face-to-face (F2F) environment when using case-based teaching. Second, the ideal case teaching method is highly contingent upon the arrangement of teaching tools (e.g., sliding blackboards to showcase divergent opinions and motivate and moderate real-time flame discussion) in the traditional environment. These tools are not available in the OAL environment. Alternative online tools (blog, discussion forums and wikis) that can substitute or complement these existing F2F teaching tools are evolving. It is crucial to assess the efficacy of these new online tools. Third, online assessment of learning performance in the case teaching method is another challenge to business faculties. Online asynchronous tools, such as whiteboard and multimedia media, allow students to present their PowerPoint slides and materials virtually. The ability of a faculty member to facilitate the constructive discussion among individuals and teams is another important factor critical to the success of case teaching method. The efficacy of case method further relies on two pillars of core concepts: (1) active learning, and (2) problem-solving learning. These two core concepts can substantially improve the learning process according to the cognitive learning theory (Alavi, 1994). Active learning is about participation. When students actively involve themselves in the analysis and discussion of cases, they immerse themselves into the role of protagonist. This can result in deepening the understanding of the studied case and increasing the urgency to resolve the case-related problems. In contrast, an



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

instructor becomes a facilitator, director, participant, and motivator, or an initiator to support the active learning of their students (Charan, 1976). In tandem with the learning process of students, the adequate facilitation of an instructor can lead students into discovery learning and high-order reasoning. The ability of students to propose practical solutions to issues embedded in the business case can be much improved. With the physical distance between instructor and students, the innovative use of online tools to promote active learning and problem solving ability of users is indispensable. The poor training of professors to conduct online courses is another threat that arises along with the introduction of online learning technology. An Education Schools Project report surveyed thousands of public school teachers, principals and superintendents, as well as faculties at universities. This report found that most graduates receiving the traditional teacher-training method are incompetent of teaching the growing diversified student populations (Williams, 2006). This calls our attention to the opportunities and threats of online case teaching method against the traditional one. An effective case teaching method needs to deliver at least four basic elements: (1) situational analysis, (2) active student involvement, (3) nontraditional instructor role, and (4) analysis-based actions (Christensen & Hansen, 1981). These elements are major forces behind the improvement of cognitive learning process. Students can effectively understand practical context in a teaching environment via these fundamental elements. The case method is often conducted more easily in the face-to-face (F2F) learning environment because of the challenges of solving tasks at the organizational level with near-toreal-life complexity. It is unlikely to find a yes or no answer for tasks of this complexity. Rather, learning outcomes are often unpredictable and vary with individuals. The case method blends four fundamental dimensions of task complexity: (1) outcome multiplicity; (2) solution scheme mul-



tiplicity; (3) conflicting interdependence; and (4) outcome uncertainty (Campell, 1988). Pedagogical solutions for these tasks need to rely on active discussion, instant feedback and problem-solving approaches. The F2F learning environment is a natural ground to exercise these pedagogies in order to help students digest cases and discover practical knowledge. Although the case method has been an important approach for professional education, the cost of conducting the case method is still high, especially for classes with a lot of students. As a result, the utilization of case method teaching can have its limitations. The online asynchronous learning environment has potentials to address these limitations. E-learning systems, such as Web-CT or Blackboard, can facilitate students’ online discussions and provide statistics on the contributions of the students. Ease of use encourages students to take part in online discussions. However, there are many key challenges to implementing e-learning systems in general. They include (1) the cost of developing and purchasing e-learning software; (2) the time required to develop e-learning courses; and (3) the need to be convinced of e-learning’s effectiveness compared to other training models (Bloom, 2003). We hypothesize that although it is feasible to teach cases in an online asynchronous learning environment, it may not be effective to promote cognitive learning processes. We conducted a study that compared the F2F and OAL environments, using the case method teaching process. This approach facilitated the discussion of a large group of students, allowed the instructor to motivate the participation of students, and did not over burden the instructor. The study was designed to provide some indicators of students’ learning performance using the efficacy of the case method teaching approach. The study also allowed the instructor to understand the problems and benefits of the online case method based on the responses of students. Finally, a pedagogical interpretation of how we can use e-learning systems effectively for teaching and learning cases is specifically posited.

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

cogniTiVE lEarning ProcEssEs The case-based teaching method has positive effects on learning because it can improve two of the cognitive learning processes: active learning and problem-solving learning. The efficacy of the case-based method while using OAL can also be measured by its effects on these two cognitive learning processes.

active learning Active learning is a goal-oriented process (Wittrock, 1978). The goal of studying cases is to understand “how” and “why” the case succeeded or failed. Since there is no single “right” answer, information acquisition alone cannot realize this goal. Case teaching method often allows learner to derive varying answers to the studied case, depending on presumptions and environmental conditions. Therefore, using a case can help a learner to not only acquire major concepts related to the studied case, but also, most importantly, its social-cultural aspects (Jones, Connolly, Gear, & Read, 2006). Different learners can reach dissimilar conclusions for each case study. Understanding a case is best accomplished by sharing mental models of learners via information exchange, analysis, manipulation, and structuring (Alavi, 1994). The case-based teaching approach can take advantage of the interpersonal interactions in a cooperative or competitive context. This practice helps activate the information processing processes between the short- and long-term memories of case study learners. Piaget (1926) and Vygotsky (1978) consider interpersonal learning an effective learning approach. In the traditional face-to-face environment, interpersonal interactions take place on a sequential basis in real time. In addition, positive or negative physical expressions (voice tones, facial expressions and body language) of discussion participants are effective enforcements for the learning process. Hence, in

the traditional F2F environment, case studies can allow an instructor to fully realize the benefits of the social learning process. Information technology has been assimilated into the classroom to enhance the active learning process. For instance, Texas A&M University Libraries are adopting an Audience Response System to assist learners in improving their active learning in the field of library instruction (Hoffmann & Goodwin, 2006). Web 2.0 technologies or conversational learning technologies, and group interactive learning technologies are emerging to cope with the challenges of delivering the active learning process in the virtual learning environment. However, in the pure online asynchronous learning environment, the dimensions of different time and different place may prevent these benefits from being achieved. For instance, in an OAL environment, instructors and students do not have to engage in the social learning process at the same time. The self-choice freedom poses two threats to the effectiveness of case-based teaching. The first is the reduced tension of interpersonal interactions that occur in either the cooperative or competitive context. Questions that need to be addressed immediately and can bring up stronger enforcements and higher-level reasoning are sacrificed with the freedom of online asynchronous learning. As a result, participants with higher self-esteem, motivation to think, or motivation to respond to discussion topics via the discussion forums or e-mails may be the main beneficiaries of higher-level reasoning. Users with lower motivation may wait to read disparate pieces of messages posted by other participants over a period of time. These students tend to wait until a few days before the deadline to post their messages. As such, the two-way interactive process between the short- and long-term memories of a learner is replaced with a passive mode of short-term memory access. The social learning process has a hard time taking place without a proper teaching policy to motivate many online learners to discuss cases. The time delay and



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

motivation issues in the OAL environment can lower the tension of interpersonal interaction, thereby reducing the effectiveness of higher level knowledge that can be realized by case-based teaching in the F2F environment. Rather than engaging students in the active learning process, the OAL environment can push students to an over reliance on the posted course materials and a delayed reading and posting of messages. Because of its passive attributes, teaching cases in an OAL environment can turn into a one-way instruction method, which is often cited as a failure of an OAL environment.

Problem-solving learning An effective business education needs to go beyond the formulation of abstract concepts, case histories and flow diagrams (Gosling & Mintzberg, 2006). Instilling users with the ability to creatively solve current and future business problems is more pertinent to the success of a quality business education. The Republic Polytechnic in Singapore adopts the problem-based learning (PBL) approach in its entrepreneurship course curriculum (Tan & Ng, 2006). PBL method is similar to interdisciplinary and “learning-bydoing” approach. This instruction method shows significant progress in enhancing students’ motivation to learn entrepreneurship and capacity for entrepreneurship. Problem solving using the case method can be impacted by task complexity. There are different views about task complexity. One valid view about the task complexity of case studies is to consider the task as behavior requirements. This view acknowledges the importance of what needs to be accomplished and how to accomplish a given task with varying characteristics (Campbell, 1988; Hackman, 1968). A case study needs to help students learn how to understand the given information, acquire relevant information (primary or secondary information), and accomplish stated goals or tasks via a critical thinking



process. The pedagogical design of case studies contains the concept of tackling tasks as behavior requirements. Tasks related to case studies can be categorized into five general types: (1) simple tasks, (2) problem tasks, (3) decision tasks, (4) judgment tasks, and (5) fuzzy tasks. These task types vary by four primary attributes that contribute to task complexity: (1) outcome multiplicity, (2) solution scheme multiplicity, (3) conflicting interdependence, and (4) solution scheme/outcome uncertainty. Since the exposure of divergent views to class members is considered by many to be a critical success factor of the case method, outcome multiplicity or more than one desired outcome of a case discussion is essential in the case study. Students can propose different strategies to attain the same goal. For instance, some students may propose cost reduction as the driver of international competitiveness while others may propose innovation. As long as the proposed solution is supported properly with the evidence and arguments, an instructor is confirmed that students have learned how to analyze a simulated case. This “solution path” is a common practice to conduct a case study. The conflicting interdependence is also commonly identified in the case-based teaching approach. Consider that both cost reduction and innovation are two feasible solutions. However, they are two potentially conflicting solutions because innovation requires slack time and human capital investments. Thus innovation-related investment often results in the increase of cost. Students can be trained to resolve the conflicting issue via case studies. After feasible solutions are located, students need to face the uncertainty challenge because many factors can affect the success of feasible solutions. For instance, sensitivity analysis and contingency planning are two general techniques that have been used to reduce the solution/scheme outcome uncertainty. In a F2F environment, an instructor can address the divergent solution paths, or can have the students analyze the differences. Task complexi-

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

ties can be discussed as they arise. Everyone hears the same arguments at the same time and in the same contextual environment. In the OAL environment, it is convenient to perform simple tasks because social presence and interactivity elements do not need to be taken into consideration. Uploading lecture slides, providing links to information on the definition of concepts or factual data, and brainstorming ideas are typical examples of simple tasks. Students can always revisit posted ideas and concepts to refresh their memory. Retaining information for longer durations increases the chance of moving the learning outcomes of simple tasks into longterm memory. In addition, discussion forums and e-mail correspondence are often employed for the communication of simple tasks with other students. Students who read the posted lessons, links, messages and e-mails may be motivated and their metacognition can be potentially improved. These communication tools may be more efficient and effective than in the F2F environment for some individuals to accomplish simple tasks. Unlike the F2F environment, competitive context among students is usually much less intense in an OAL environment because no live debate actually takes place. Live debate is replaced by the self-paced reading of posted messages and voluntary responses to different opinions. The many-to-many relationship among class members in a F2F environment is no longer an indispensable element in an OAL environment because communication often takes place in one-to-one or one-to-many relationships. Furthermore, the social presence element is not a part of the OAL environmental communication process. Although online learners are also exposed to similar and divergent views of students, the cardinality difference and low social presence may not have the same force on the mental models of students. The increased fluency of ideas is usually not translated into qualitative ideas in an OAL environment. In this environment, students may not benefit by the extension of their mental model and meta-

cognition. Along with the delayed response, it is difficult to address most questions in a timely manner (unless the professor or another student just happens to be online at the time of the question). The decreased group support may have two side effects. The first is to lower the satisfaction level of some online learners, thereby resulting in a lower intention to continue usage. Second, some learners can suffer from lower motivation, which could lead to lower cognitive gains. The more complex and conflicting the tasks (such as decision tasks, judgment tasks, and fuzzy tasks) that need to be resolved in an OAL environment, the less effective learning outcomes can be compared to the F2F environment. Using the case method to teach in the OAL environment may not be as effective as in the F2F environment because many complex tasks can be involved. Key elements such as a larger class size, information overload, incompatibility with certain tasks, and weak support for active learning may make OAL a great learning environment to motivate students to learn, but not necessarily to achieve a higher order of cognitive gains. The motivation to participate in the online discussion can be measured by the number and quality of posted messages and feedbacks. In contrast, the cognitive gains may be assessed with a student’s performance in exams.

TEaching ProcEDUrEs oF a casE mEThoD class Charan (1976) suggested that the success of the case method depends on three sequential steps: (1) course design, (2) class preparation, and (3) method of conducting class discussions. We concede that course design and class preparation are vital steps in the case method of teaching and are basically carried on by the professor in the same way for F2F and OAL environments. We even believe that the way F2F and OAL classes may be conducted may be very similar. Where we think that there



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

is a divergence is in the ability to develop higher order reasoning in F2F and OAL classes. We propose that, normally, an instructor follows four general phases to conduct case method classes in both the F2F and OAL environments (Figure 1). While we recognize that there may be some differences among professors, we believe that each of these four phases is usually addressed, either implicitly or explicitly. Distinct differences between teaching cases in F2F and OAL environments emanate when being examined along casebased teaching procedures. Two of the differences may be the use of time and the communications media used for case analysis.

Phase 1. introduction of concepts The instructor begins with an introduction of the concepts to be covered in the case. Different instructors may have different emphases and different methods for presenting concepts. In the F2F environment, emphasis can be stressed with facial expressions or voice tones of instructors. In addition, students can ask questions and get feedback if they do not understand the expectations of the instructor. Based on the instant feedback of students, the instructor can immediately solicit useful information to modify or adjust his original teaching plan according to the responses of the students. In the OAL environment, students control their own learning pace because the voice and facial expressions of instructors are substituted with anything from text-based instruction to prepared streaming videos or video downloads. Although students can send an instructor e-mails asking questions, the feedback process tends to be delayed. Therefore, questions regarding the concepts covered in the case often take place when students encounter difficulty with assignments.



Phase 2. student case analysis In this phase, each student team must analyze the case assignment. Energy committed activities, such as information gathering and preparation for individual parts of the case, are required to properly prepare an analysis. Whether students perform the case analysis in class or outside of class, the F2F environment provides a much richer environment for case analysis. In the F2F class, students might sit close to the instructor or with other students so they can discuss the particulars of the case. Divergent views are readily available in the in-class discussion. The internalization and externalization of the knowledge management process can be easily acquired in F2F. This can advance learners into a higher level of reasoning because communication centers on structuring information processes and processing acquired information. Information shared in class is available to all students. In the OAL environment, the learning environment is usually more formal and less competitive because students never see each other; only the output is seen. There is a physical and psychological distance in this environment, which can distract online learners from direct attention to the case analysis. Since the mutual reciprocity and competition is usually lower in the OAL environment, it may not provide as good a motivational learning environment in which students should extend their metacognitive ability. Students may be able to see both relevant and irrelevant information available on the discussion board or threaded discussion. Online learners can easily copy information from the Internet and suggested Web site links to address other students’ questions. To summarize, it is therefore possible that the F2F environment is more suitable for highlevel reasoning while the OAL environment is more suitable for low-level reasoning in the case analysis activities.

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

Figure 1. General phases to conduct case method classes F2F • High Social Presence • High Information Richness • Instructor control • Instant interpersonal interaction Higher motivation Cooperative content Physical closeness Psychological closeness • Informal student interaction • • • •

• Active feedbacks • Lower exposure to divergent views • Higher-level reasoning & critical thinking • Active discussions

• Oral and written feedback • Immediate gratification

Phase

OAL

Introduction of Concepts

• Low Social Presence • Low Information Richness • Learner control • Delayed and passive interaction

Student Case Analysis

Output Generation and Discussion

Follow-up and Evaluation

Phase 3. output generation and Discussions Usually the output is a report that presents the results of the case analysis, either oral or written. In the F2F environment, output can take the form of many things: written report, PowerPoint presentation, oral presentation, or the student leading the class discussion. If an oral presentation is made or the student leads the class discussion, then the higher order reasoning or problem solving results can be presented and defended in a relatively short period of time. In this environment, points can

Lower motivation Competitive content Physical distance Psychological distance • Formal student interaction • • • •

• Passive feedbacks • Higher exposure to divergent views • Lower-level reasoning & critical thinking • Passive discussions • Written feedback • Delayed gratification

be made, questions can be asked, and counterreasoning can be presented while the ideas are fresh, thus allowing students to become active in their questions and counter analyses. Both the cooperative spirit of the students and the competitive spirit among students can blossom. In the OAL environment, output generation is somewhat limited to a written report, sometimes accompanied by a PowerPoint presentation. Students communicate with the presenting student by e-mail or discussion tool. Questions and answers can drag over long periods of time. This can be good – it allows students to think about and



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

analyze the points presented and develops richer questions and counter analyses – or bad – students lose interest in the ideas and become passive in their approach to the presentation.

Phase 4. Follow-Up and Evaluations Follow-up and evaluation is the final phase of the process. In a F2F class, the instructor can provide students with immediate feedback. This usually takes place with class discussions about the salient points made in the case analysis. Areas that might have been omitted or overlooked can be explored. Specific points that the instructor wants to cover can be presented and are heard by all students. In a F2F class, the instructor must be very prepared and quick to analyze the analysis and presentation to cover all of the points. Individual evaluations usually take place in the form of written feedback. Follow-up and feedback in the OAL environment is almost exclusively written, delayed, and passive in nature. The instructor usually has more time to read, evaluate, and respond to each student’s analysis. However, in this environment, the other students in the class benefit less from the follow-up because they may not be privy to the written communications (e.g., grading comments). Because the feedback and evaluations are written, this phase may take more time on the instructor’s part than in a F2F environment and can be less effective.

rEsEarch DEsign An experiment was conducted in a field setting that enabled the study to garner greater external validity than would be the case with a laboratory experiment. A field experiment methodology has the merits of “testing theory” and “obtaining answers to practical questions” (Kerlinger & Lee, 2000, p. 583). The exploratory nature of the study requires that variables (e.g., interaction



modes and usage patterns) under investigation be carefully observed and interpreted. As such, the experimental study is combined with the content analysis technique of the qualitative research method to verify and/or modify our statistical interpretation. The setting for the field experiment is a “Management Information Systems” course offered by one university in Taiwan to 124 senior undergraduates of two classes who majored in business administration. The instructor, course materials, learning content and evaluation criteria were all the same in the two classes. The case teaching method in the experiment combined F2F and OAL, tried to facilitate the discussion of a large group of students, and tried to motivate the participation of students.

hypotheses This study focused on investigating the impacts of course designs (F2F vs. OAL environments) on learning outcomes. The following hypotheses were proposed for this study: H1: The number of posted messages in the online discussion sessions is positively correlated with a student’s graded performance in case discussion assignment. H2: The number of feedback messages in the online discussion sessions is positively correlated with a student’s graded performance in case discussion assignment. H3: The number of posted messages in the online discussion sessions does not contribute to cognitive gains. H4: The number of feedback messages in the online discussion sessions does not contribute to cognitive gains.

Experimental Design Students are graded based on two assignments: case reports for the team and an online case dis-

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

cussion. Students discussed a total of five cases. There were three questions in each case for students to discuss over a period of two weeks. Each case was taken from Laudon and Laudon’s (2002) translated textbook Management Information Systems. The instructor decided the topics for each case discussion. Students were required to use the same online asynchronous learning tool developed by Xoops (http://www.xoops.org). Therefore, they are exposed to the same online discussion interface. This interface is very much like most of other online discussion interfaces (e.g., WebCT or Blackboard). The system also provides a test area for novice users and chatting boards to improve student’s motivation to participate. Students were required to adhere to the procedures described in the next subsection when using the online discussion forum.

Procedures The experiment’s procedure utilized in this study followed the stages in Figure 1 and combined the online discussion, in class F2F activities, and assignments for the student to accomplish on their own. The procedure and timeline are outlined and described below.

Phase . Introduction of Concepts The instructor announced the discussion topic that was to be discussed in each two-week period and introduced the relevant concepts to be explored. Specific questions to be addressed were assigned to teams of students. Each team consisted of four to five students.

Phase . Student Case Analysis In the first week, teams analyzed the case and prepared a team report. They were required to publish their answers to the discussion topics.

There were a total of three questions for each case, and 4 teams for each question.

Phase 3. Output Generation and Discussion During the second week, students could critique or modify the posted discussion results, or suggest other alternatives on an individual basis. Students could also raise new questions or viewpoints regarding the discussed case. Each team was supposed to address the individual feedback of other students in the class to their posted discussion results. Each online discussion session concluded at the end of the second week. Teams were required to make a presentation in class regarding their discussion dynamics, such as the posted messages, questions, discussion procedures, discussion results, and contributions of students. In addition, students were also required to compare their answers with other students’ answers. As a final output, each team was to submit a written report for evaluation purposes.

Phase . Follow-Up and Evaluation A student’s final grade was based on two assignments: case reports for the team and an individual final report. The case report is comprised of a written report and a presentation. Additionally, each student was required to organize the discussion threads into an individual written discussion report and submit it to the instructors for grading in the end of the semester. The final discussion report was to include an index page summarizing the number of posted messages and feedback, and all the posted messages behind. The case report and individual discussion report accounted for twenty percent of a student’s total final grade. To improve the learning effectiveness and mitigate the potential problems raised previously, students were asked to adhere to the following rules:



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

•

•

•

• •

•

The number of posted messages is not part of the final grade. Rather, only the number of posted messages considered as useful or influential to other students’ learning is graded. Therefore, any messages that not make a contribution to the online discussion session as a whole received a lower score. Students making no comments in the online discussion will not receive individual scores. Each opinion needed to have clear objectives. For instance, students need to say that “Question,” “Propose for Modification,” “Explanation,” “Propose New Solutions,” and so on. Students were also encouraged to take a stand on “for” or “against” a specific comment. Comments supported with references or a URL were encouraged. Students were advised of the hierarchical structure of the posted messages. When making a comment, students needed to specifically point out which messages that they are addressing to avoid confusion. Although sometimes they had questions about students’ comments, the instructor

deliberately did not intervene and interrupt the online discussion sessions. At the end of each case discussion, the instructor explained the purpose of questions and comments on proposed answers towards those questions. The research design was developed to address the hypotheses using specific surrogate measures in an online asynchronous learning environment.

DaTa analYsis anD FinDings A total of 925 messages were collected from the discussion board, testing areas, and chatting boards at the end of the five cases in the discussion. The number of messages affirms the ability of e-learning systems to produce a large quantity of messages. Exhibit 1 shows that the number of messages posted in different online discussion sessions. The number of posted messages varies greatly in different weeks. This seems to indicate that many students waited until the end of the semester to start joining the online discussion

Exhibit 1. The number of messages posted in different discussion sessions Submission Prepare Report In-class Discuss Case  - Week  Case  - Week  Case  - Week  Case  - Week  Case  - Week  Case  - Week  Mid-Term Case  - Week  Case  - Week  Case  - Week  Case  - Week  Orientation 0

0

00

0

00

Number of Messages

0

00

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

sessions to make up their scores. The indicator corresponds with the small number of messages posted in the preparation week before the formal discussion sessions were conducted. These messages were captured in the testing area and chatting boards in the preparation week. The number of posted messages is a proper surrogate to assess the motivation of students to participate the online discussion sessions. In the beginning, students were attracted to the collaborative online learning approach. This is reflected by the increased number of messages posted in the discussion board. However, the number of posted messages decreases in the second case study compared to the first case study. This seems to indicate a lower degree of motivation. This may be caused by the curriculum design or factors associated with the second case being studied (i.e., complexity of the case, interest of students in the company or topic, etc.). After the mid-term, students seemed to be paying more attention to their grade, possibly because of the influence of their mid-term grade. This resulted in an increased number of posted messages. Participation rate in general gradually improved. The participation rate further improved as the schedule got closer

to the end of the semester. The number of posted messages reached the highest at 254 in the final week of online discussion. It should be noted that all case discussions were open for the entire semester to motivate learners to participate the online discussion. Students could discuss earlier cases even before the end of the semester. Since the participation rate was very low in the first and second case discussion sessions, students were encouraged to add to their discussion of earlier cases after the fifth case was concluded to help them improve their grades. The final grade was based on the messages posted up to half of the case report preparation week. A total of 41 messages were posted after the fifth case discussion session was concluded. These messages were counted and helped those participants improve their individual grade. A total of 124 students registered for the Management Information Systems class. In the end of the semester, six students failed the class. Exhibit 2 shows the number of messages posted by individual students. A cursory investigation of Exhibit 2 shows that most of the students posted fewer than 15 messages. There are only eight students who posted more than 15 messages and

number of Users

Exhibit 2. The number of messages posted by individual students     0     0 



 0          0 number of messages



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

their messages are scattered irregularly. Among these students, the largest majority of students contribute to five cases. If the total number of messages is divided by five cases, there is only one message per case per student. This finding indicates that most of the students are not motivated to learn via the online discussion forum. It is also hard to imply that the interaction among students takes place in the online discussion based on the findings. We further analyzed the correlation between the online participation in the discussion forum and individual learning outcomes. Each variable is defined as follows: •

•

•

•



The number of posted messages (PM): The OAL system automatically records the posted messages and can calculate the number of total messages posted by students. This number is a surrogate to evaluate a student’s participation rate. The number of feedback messages (FM): When students respond to a posted message, the OAL system automatically records the message and totals the number of messages. If the posted message has value to the discussed topics, a student can often solicit more feedback than from the posted messages that have less value. Therefore, the number of feedback messages is considered as a better surrogate than the number of posted messages to reflect the value of posted messages and their contribution to the online discussion sessions. Case discussion (CD): Instructors graded each individual’s term paper based on their subjective evaluation of the following criteria: the number of posted messages, content, contribution of messages to other students, triggers for interactive discussion, and time commitment. The total grade (TG): The final grade of the semester, a total grade of individual case discussions, mid-term and final exams, and assignments.

•

• • •

The non-case discussion grade (NCDG): The deduction of individual case discussions from the total grade. The grade represents the performance that is not related to individual case discussion. Test scores (TS): The total of mid-term and final exam scores. Case reports (CR): Team’s written report and in-class presentation. Assignment Scores (AS): The total score of assignments.

Table 1 shows a correlation among test results. Most variables of this study are correlated with each other except those variables in the left bottom corner. The rest of correlation coefficients are significant at α=0.05 or α =0.01 level. These findings indicate that among the tested variables the number of posted messages, the number of feedback messages, and case discussion scores are highly correlated with each other. These variables belong to one group. These results seem to support Hypotheses 1 and 2. In contrast, the non-case discussion grade, test scores, case report and assignment scores are correlated with each other and belonged to another group. These results seem to support Hypotheses 3 and 4. Since the final grade is comprised of these two groups of variables, it is correlated with all variables. No significant correlation coefficients can be detected between these two groups of variables. There possibly exists an orthogonal relationship. The online case discussion grade has little relationship with learning performance. So if those indicators can be the surrogates of learning performance, the online case discussion may not be an effective approach. Our observation suggests that given the limited discussion topics and contents in the online learning environment, the online case discussion may not greatly help improve an individual’s ability to solve case-related problems. The grade of the online discussion is primarily affected by a student’s degree of participation. Increasing the degree of participation might be

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

Table 1. The correlation analysis of dependent variables PM PM

FM

CD

TG

NCDG

TS

CR

AS

1.00

FM

0.86 ** p=0.00

1.00

CD

0.67 ** p=0.00

0.61** p=0.00

1.00

TG

0.21* p=0.02

0.22* p=0.02

0.40** p=0.00

1.00

NCDG

0.02 p=0.81

0.05 p=0.59

0.13 p=0.16

0.96** p=0.00

1.00

TS

0.00 p=0.97

0.07 p=0.43

0.10 p=0.26

0.94** p=0.00

0.98** p=0.00

1.00

CR

0.07 p=0.44

0.14 p=0.14

0.16 p=0.08

0.47** p=0.00

0.46** p=0.00

0.31** p=0.00

1.00

AS

0.07 p=0.44

-0.02 p=0.86

0.17 0.06

0.45** p=0.00

0.44* p=0.00

0.32** p=0.00

0.30** p=0.00

1.00

** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

easier than improving the grades in the mid-term or final exams for some students to improve their final grade. This rationale can be justified because the online case discussion can properly reflect the commitment of a student to the discussion topics, but not necessarily to the learning outcomes. With this logic in mind, it may be important to integrate the online discussion into a curriculum in order to motivate students to learn. There are three additional findings to report. The grade of the case reports was correlated with learning outcomes, but not with grades in the individual case discussion. Secondly, the scattered distribution of the number of posted messages suggests that it may be easier to improve an individual’s learning outcomes by having a lengthy and more organized discussion. Thirdly, although we asserted that the number of posted messages and the number of feedback messages might mean different things; lower correlation coefficient between these two variables suggests otherwise. When grading individual case discus-

sions based on the listed criteria, the instructor’s judgment is somewhat influenced by the number of posted messages. The high correlation between individual discussion grades and the number of posted messages indicates along that line.

limiTaTions Our hypotheses were built upon the general assertion that the motivation of students is positively correlated with their performance. This assertion has been supported in the literature of end-user computing, educational psychology, and so on. However, when applying the assertion to this study’s hypotheses, we may need to be more sensitive to the context of learning using the case method in the OAL environment. For instance, the number of posted messages may not necessarily be an accurate surrogate to measure the motivation of students. It will take some time before more objective metrics can be developed



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

to accurately measure the online learning performance of students. This is similar to evaluating the effectiveness of online advertising based on the click rate alone. Nowadays, search portals, such as Google, are developing better metrics, such as actual click-through-rate (CTR) and cost-per-click (CPC) to reflect the accuracy of searching results. Thus, it may not be appropriate to propose that the number of posted messages can contribute to the cognitive gains of students. One salient indicator is that most messages were posted right before the end of the semester. Learning is a timeconsuming process, particularly in business education. Many intangible learning processes (e.g., social interaction, verbal jousting, and the friction between collaboration, leadership, creativity and competitiveness) take place, in addition to the studied subjects (Bradshaw, 2006). The efficacy of case teaching method can be highly discounted if an appropriate online assessment is not well developed to capture these intangible elements. Many researching opportunities continue to exist in this online assessment area. Straub (1994) suggests that people from different cultures, particularly high vs. low individualism, have different perceived usefulness and ease-of-use of computer-aided communication and training. For computer-aided communication particularly, IS studies show that individualism is a good indicator to measure and understand usage behaviors of information systems. The individualism tends to influence end-user in the opposite direction from the other three cultural dimensions – power distance, uncertainty avoidance and masculinity. This finding has been proved in many IS studies with many information systems, including Internet usage (Straub, 1994; Straub, Keil, & Brenner, 1997; Moon & Kim, 2001; Lederer et al., 2000). Culture’s influences can also be found in the OAL environment (Chu, 1999; Svastisinha, 1999). For example, the design of a home page or the OAL interface (a Chinese interface called Xoops was used in this experi-



ment) can impact a study such as this. The subjects of this study have the same cultural background where individualism is perceived as low. This may verify with an overall low motivation of users of this study to learn the case method in the OAL environment. There are other pedagogy methods and presentation media that can be used in the OAL environment. For instance, the media with higher social presence effect, such as video-on-demand, audio-on-demand, three- and four-dimensional interactive media, and text-to-audio conversion software, may be a way to improve end-user satisfaction (Short, Williams, & Christie, 1976; Yoo & Alavi, 2001). The simulation software to guide students in solving problems gives students a higher degree of control of their learning path and speed. Interactivity is usually considered as an important design factor for online programs. Lavooy and Newlin (2003) asserted that the effectiveness of online courses is highly correlated with the interactivity levels of online learning systems. Thus, learning the case method via the simulation software may improve students’ ability to solve problems, thereby improving their cognitive gains. Real estate, automobile and apparel companies are now offering 3D online touring feature on their Web sites to increase the degree of information transparency and customization. Another variation of simulation software is the use of online learning games based on Macromedia Flash, to improve learning effectiveness (Huang & Cappel, 2005). Not only can their users advance their knowledge on prospective products and services, but also shorten the time to make decision. Web 2.0 technologies are emerging to empower users to converse and socialize with each other to improve communication capacity. A natural learning experience is a social process. Most elearning designs trade technology for the social learning process. Online learning will lose its effectiveness if this core learning element is missing (Cardus, 2006). People primarily adopt these

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

Web 2.0 technologies for socialization purpose (e.g., myspace.com). It requires a major mindset shift and more pedagogical supports before we begin seeing people leverage Web 2.0 technology as a case-based learning tool. However, this study did not investigate the potential effects of social learning technology.

conclUsion The study identified some interesting conclusions. First, the OAL environment can promote students’ in-class participation under certain circumstances. In the F2F environment, students sometimes do not attend classes, thus losing their participation. However, in the OAL environment, students can not just “miss class” because class is always in session. The OAL environment has logs to keep track of students’ activities. This mechanism can motivate students to participate. Second, students have difficulty resolving complex and conceptual problems in the OAL environment. Despite this, OAL is a fertile ground to resolve clearly defined problems, such as data collection, exemplars, and self-introduction. Third, learning how to organize thoughts and write in a succinct way to help communicate is another important accomplishment. This may not be the case in a spontaneous F2F class environment. For example, in a F2F class, short answer or essay responses to exam questions are seen by only two people - the student and the instructor. In an online class, if students could post their short answer or essay responses to exam questions on the discussion board for everyone to read, they might be more careful about their responses since every student could read the answers and challenge them. The grade in the online discussion activity has little correlation with the indicators of learning outcomes. This indicates that the OAL environment may not directly contribute to learning outcomes as it does to motivation. From the theory of constructivist learning, it may be important to

adopt the hybrid approach in the OAL environment, such as using an online discussion forum to improve participation rate, but clarifying ambiguous concepts and topics via by a more personal approach, such as e-mail. Harvard and Stanford, two premier business schools, are collaboratively offering such a hybrid Executive MBA program in a CD package format (Bradshaw, 2006). Finally, the assessment of learning outcomes could also vary with the objective of online learning classes. The metrics to measure the efficacy of case teaching method is multifaceted, including qualitative and quantitative analytical skills, participation, reflection, creating new ideas, leadership, and so on. An instruction needs to continuously explore better assessment tools to measure learning objectives in the online environment. The salient information overload problem caused by the OAL environment needs to be carefully addressed with a proper curriculum design. One feasible solution is to structure the online dialog and documents (Turoff et al., 1999). Another alternative is to train students with the basic inference and logical thinking ability so as to promote the effective exchange of ideas. Indeed, there are different approaches to utilizing the case method in the OAL environment. These options never stop evolving since Web technologies are continuously being developed. Cases in point, Google attempts to digitalize all the library resources of the world in order to improve the efficiency and effectiveness of searching process for scholarly literature. Its current service offering (http://scholar.google.com/) is a precursor to this grand vision. Wikipedia is another major force flattening the world (Friedman, 2006). The interplay between technological capacity and creative instruction methods could foster opportunities in the area of case teaching method in OAL environments. Further research may need to better assess the impacts of different online learning approaches.



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

FUTUrE rEsEarch The data obtained from this study partially supported its expected relationships between variables. It is possible that a more accurate surrogate is needed. Subsequent research is need to establish the validity and reliability of different surrogates to measure the satisfaction level and learning performance of students to learn using the case method in an OAL environment. For instance, the frequency of posting messages, the scope of discussion, and the relevance of posted messages towards to the discussed topics are areas needing further research. In a F2F class, students are not able to go back and discuss past cases to improve their grades. Therefore, the ability to post messages to previous case discussions should be examined in future research. In addition to the possibility that the number of posted messages and feedback are not adequate for the purpose, there is always the possibility that the other independent variables could have been inadequate as a representation of its intended constructs. It may be important to conduct a study to understand the perceived ease-of-use of the e-learning system used by the students. The way the posted messages are formatted and organized can influence students’ motivation to use the e-learning system. In the class used for this study, students complained about the difficulty of reading other students’ comments via the default hierarchical structure in the class management software. Different ways (different color, rankings, currency of comments, etc.) to organize the posted messages may be a factor that needs to be further explored. With any complex treatment, such as the establishment of online discussions, there is a chance that operationalization of the variables can be less than what is needed for effects to occur. Additional research is required to refine and perfect the online discussion treatments as much as possible. There is no simple manipulation check for verifying the efficacy of this kind



of treatment, but continued investigation should reveal the extent to which the manipulation is a successful one. Future research can attempt to improve the reliability of these findings by controlling the experimental environment more tightly (e.g., alignment between the performance surrogates and grading policy) or by improving the strategy’s generalizability through the examination of other variables (e.g., students vs. professional workers, number and duration of training sessions, type of training media, self-efficacy, experiences of using the online learning system, and software types). With the increased dependence on distance learning in the educational field, additional research needs to be performed on its effectiveness and how to improve on its efficiency. We hope others will look more closely at the OAL environment as a fertile research area. This could significantly add to the quality of distance education (online asynchronous learning) in the future.

rEFErEncEs Alavi, M. (1994). Computer-mediated collaborative learning: An empirical evaluation. MIS Quarterly, 18(2), 159-174. Bloom, M. (2003). E-learning in Canada: Findings from 2003 e-survey. Bradshaw, D. (2006, December 11). An MBA is what you make of IT. Financial Times, pp. 14 Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 13(1), 40-52. Cardus, J. (2006, November). The power of the collective. E.learning Age, pp. 10-12. Charan, R. (1976). Classroom techniques in teaching by the case method. The Academy of Management Review, 1(3), 116-123.

Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

Christensen, R., & Hansen, A. (1981). Teaching and the case method. Boston: Harvard Business School Press. Chu, G. L. (1999). The relationships between cultural differences among American and Chinese university students and the design of personal pages on the World Wide Web. Unpublished doctoral dissertation, University of Georgia.

Lavooy, M. J., & Newlin, M. H. (2003). Computer mediated communication: Online instruction and interactivity. Journal of Interactive Learning Research, 14(2), 157-165. Lederer, A. L., Maupin, D. J., Maupin, D. J., Sena, M.P., & Zhuang, Y. (2000). The technology acceptance model and the World Wide Web. Decision Support Systems, (29), 269-282.

Friedman, T. L. (2005). The world is flat: A brief history of the twenty-first century. New York: Farrar, Straus, & Giroux.

Moon, J., & Kim, Y. (2001). Extending the TAM for a World Wide Web context. Information & Management, (38), 217-230.

Gosling, J., & Mintzberg, H. (2006). Management education as if both matter. Management Learning, 37(4), 419-428

Piaget, J. (1926). The language and thought of the child. London: Routledge & Kegan Paul.

Hackman, J. R. (1968). Effects of task characteristics on group products. Journal of Experimental Social Psychology, 4(2), 162-187. Hoffman, C., & Goodwin, S. (2006). A clicker for your thoughts: Technology for active learning. New Library World, 107(9/10), 422 Huang, Z., & Cappel, J. R. (2005). Assessment of a Web-based learning game in an information systems course. The Journal of Computer Information Systems, 45(4), 42-49 International Data Corporation. (2002, September 30). While corporate training markets will not live up to earlier forecasts, IDC suggests reasons for optimism, particularly eLearning. International Data Corporation. Retrieved March 5, 2003, from http://www.idc.com/getdoc. jhtml?containerId=pr2002_09_17_150550 Jones, C., Connolly, M., Gear, A., & Read, M. (2006). Collaborative learning with group interactive technology: A case study with postgraduate students. Management Learning, 37(3), 377-396 Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral research. New York: Harcourt. Laudon, K. C., & Laudon, J.P. (2002). Management information systems. Prentice-Hall.

Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. New York: Wiley. Straub, D. W. (1994). The effect of culture on IT diffusion: E-mail and FAX in Japan and the U.S. Information Systems Research, 5(1), 23-47. Straub, D., Keil, M., & Brenner, W. (1997). Testing the technology acceptance model across cultures: A three country study. Information and Management, 33, 1-11. Svastisinha, R. W. (1999). Wahhn: Web-based design. Wind and human comfort for Thailand. Unpublished doctoral dissertation, University of Southern California. Tan, S.S., & Ng, F. (2006). A problem-based learning approach to entrepreneur-ship education. Education & Training, 48(6), 416 Turoff, M., Hiltz, S. R., Bieber, M., Fjermestad, J., & Rana, A. (1999). Structures in computer mediated group communications. Journal of Computer Mediated Communication, 4(4). U.S. News. (2003). E-learning: Online graduate degrees. Retrieved November 18, 2003, from http://www.usnews.com/ usnews/edu/elearning/ directory/gradonline.htm



Challenges in Delivering Case-Based Teaching in the Online Asynchronous Learning Environment

U.S. News. (2007). E-learning guide. Retrieved January 10, 2007, from http://www.usnews.com/ usnews/edu/elearning/ tables/mba_reg.htm

Williams, M. R. (2006, September 19). Study says teachers are poorly trained. Knight Ridder Tribune Business News, p. 1

Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Wittrock, M. C. (1978). The cognitive movement in instruction. Educational Psychologist, 13, 15-29.

Wang, H-W. (2006). The implementation of an asynchronous Web-based gaming & simulation environment and its effects on financial engineering education. Journal of American Academy of Business, 9(2), 360-369.

Witzel, M. (2005, October 7). A new era for digital learning. Financial Times, p. 2



Yoo, Y., & Alavi, M. (2001). Media and group cohesion: Relative influences on social presence, task participation, and group consensus. MIS Quarterly, 25(3), 371-390.



Chapter IV

Motivation-to-E-Learn:

A Quantitative Design Technique M. A. Rentroia-Bonito Technical University of Lisbon, Portugal J. A. Jorge Technical University of Lisbon, Portugal C. Ghaoui John Moores University, UK

absTracT One of e-learning challenges is to promote effectiveness in order to fully get expected benefits. Achieving effectiveness will contribute to its establishment as a credible way to support educational endeavours. To address this complex and multidisciplinary challenge, development teams need proper design techniques to build effective learning experiences. The literature does not show solid quantitative approaches to support learning-centered design, where student needs and their immediate and broader contexts are taken into account. This work explores a variable called “motivation-to-e-learn,” a key component to design technology-supported learning experiences. Our goal is to identify what motivation-related variables are critical for student engagement in learning online. This will be the basis for a specific, bottom-up and quantitative design technique. To this end, we further explored the importance of a set of motivation-to-e-learn variables building on previous results in real instructional settings. From this activity, an exploratory two-factor structure emerged which explains 96% of motivation to e-learn construct. We discuss our results, together with their implications for learning-support design and future work. Our contribution is a step towards quantitatively understanding and cost-effectively improving the link among learning-design process, supporting systems and students into an effective and harmonious whole.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Motivation-to-E-Learn: A Quantitative Design Technique

inTroDUcTion Higher education institutions are concerned with technology-assisted learning experiences, since expected benefits are yet to come. Some institutions have implemented blended-learning initiatives to test the concept within their boundaries, though this has been done in a fragmentary manner with partial results and without proper monitoring mechanisms (Britain & Liber, 2005; O’Neill, Singh, & O’Donoghue, 2004). Consequently, building cost-effective and satisfying learning experiences supported by technology is still an art, and many factors are involved. Development teams are challenged to look at technology-supported learning in a holistic manner, taking not only usability of learning-support systems into account, but also context, content, and individual-related variables. This demands integrated approaches to evaluate. To this end, they must address the whole set of specificities by anticipating breakdowns or factors that can impair learner experience, specifically their motivation to learn and perform online. Motivation is an individual variable, which is both shaped by internal factors and influenced by context (Organ & Bateman, 1991). It captures the impact of events that may take place within the individual’s immediate context. Such events may affect student efforts and results. Understanding what aspects could affect motivation-to-e-learn the most is critical to improve acceptance behaviours, and consequently to increase learning effectiveness. To this end, bottom-up and quantitative design techniques are needed to measure and improve learning-related aspects that can influence student motivation to learn. However, motivation to learn in a technologysupported learning context is difficult to measure and capture. This chapter’s goal is to propose a design technique to capture potential motivating items that can guide development teams’ efforts on improving technology-supported learning experi-

0

ences. To this end, we build on previous results and further analyse the underlying factors of this hypothetical variable by revising its identified content and using a larger student sample than that used in previous studies. Our main contribution here is to offer development teams a holistic view and a quantitative design technique to assist them in the adoption of technology-supported learning. In the remainder of this chapter, we present our conceptual framework, the study methodology, the results garnered and discussion and finally a general conclusion and future work.

concEPTUal FramEWorK The complexities associated to technology-assisted learning require holistic views that address context, process, system, and individual-related specificities. These holistic views could be supported by learning-centered design, which focus on students, instructors, institutions, and society needs (Andersen, 2004; Costabile et al., 2006). As can be seen in Figure 1, at operational level, tasks link learning-design processes to supporting systems. Learning tasks are performed by students to achieve goals. Further, the functional and non-functional characteristics of supporting systems reflect strategic and design options regarding skills and system development. The structure of the experience is given by the learning-design process and supporting system. This processsystem fit influences the quality of interactions between students and systems thus influencing their motivation to engage and perform learning tasks (Walker, 1992; Organ & Bateman, 1991; Gagne & Deci, 2005). That is why a better understanding of motivation to e-learn could help development teams to structurally design costeffective learning experiences. The scope of this study is an individual-related hypothetical variable or construct called motivation-to-e-learn. In particular, we aim at exploring its underlying nature and its implications for de-

Motivation-to-E-Learn: A Quantitative Design Technique

Figure 1. Conceptual framework Context Learning-design P rocess Role

Learner

Task

Interaction

S upporting system

Learning results

sign and improvement. Towards this end, we define motivation-to-e-learn as an individual construct denoting internal set of processes (both cognitive and behavioral) by which human energy becomes focused on learning particular content to achieve specific learning goals. This study was concerned with providing designers with a quantitative design technique that allow them to give students reasons to focus their energy on learning online. This construct addresses students’ most important concerns, including context, content, system, and personal beliefs about e-learning. Indeed, this construct summarizes a set of context-specific and internal reasons from the conceptual model of any technology-supported learning experience that extrinsically or intrinsically could be motivating for students. The content of this hypothetical variable was identified in the literature (Bandura, 1997; Costabile et al., 2005; Organ & Bateman, 1991; Walker, 1992, 2003; Wentling et al., 2000), and also from observing and interviewing users. Resulting content was tested with students in real instructional settings along the last three years. This construct is composed of twenty-two potential items centering design into the whole learning process within instructional settings. That is why the motivation-to-e-learn construct can be used as a target for improvements, and also as a communication vehicle, within multidisciplinary

development teams, to standardize understanding about student concerns to learn online.

sTUDY mEThoDologY Based on our previous preliminary results, we were interested in fine-tuning our proposed scale and determine a few important common factors underlying it. In doing so, we expect that analysing these factors would make improvement process easier and cost-effective. To achieve this study goal, a research group from our university instantiated the operational level of our conceptual framework into a onesemester multimedia course. This course was lectured in two campi and 143 students registered for this course. Learning content was previously produced in different formats (.pdf, .zip, and Webcast videos) and delivered to registered students by using an adapted supporting system. To simulate a classroom scenario, this system includes a learning management system and a Webcast component. Its main functionalities supported defined learning tasks, the need for online tutoring, and student skills. Figure 2 shows a high-level view of system architecture. In order to achieve the goal of this study, an online questionnaire was developed and tested.



Motivation-to-E-Learn: A Quantitative Design Technique

Figure 2. High-level view on system architecture

This questionnaire included twenty-two items and used a Likert-type six-point rating scale to facilitate distinguishing between positive and negative opinions, thus making data interpretation easier (Oppenheim, 2001) and contributing to scale reliability (Nunnally, 1978). The questionnaire was peer-reviewed and tested several times during the last three years. Improvements were made based on student feedback. Each item has a long description to uniform participant’s understanding on each. Also, each point in the rating scale has a label to avoid confusion about its meaning, ranging from “Of no importance to me” and “Of high importance to me.” In this study, we used the fourth version of online questionnaire. The questionnaire in its front page provided basic information about this study goal and instructions. Both anonymity and confidentiality were stressed and assured. Students were informed at the beginning of the semester about this study goal and dynamics. Students were asked to fill out the questionnaire, which was available on the system, during the second week of the semester. 71% of them timely participated. Based on system data, on average, they spent around five (5) minutes to answer the questionnaire (Rentroia-Bonito & Jorge, 2004; Rentroia-Bonito et al., 2006).



Of the participating 102 students, 76% were male, 57% registered at Campus A; 85% were between 20-24 years old, 16% held partial jobs. 92% accessed Internet at least 2 hours/day and 89% did it at more than 512 Kbps. 28% of participating students were majoring in multimedia and intelligent systems area, 84% had never participated in a similar online learning experience, and all had their own personal computer to support learning tasks.

rEsUlTs The underlying assumption of this study was that motivation-to-e-learn consists of common factors. These common factors are responsible for the observed correlations among their respective items (Johnson & Wichern, 1992). That is why we used factor analysis to process this data and achieve this study goal. Before doing so, an internal reliability analysis was performed to identify items that little contribute to the scale reliability (Oppenheim, 2001). Seven variables were eliminated: “Adequacy between content and learning objectives,” “Convenience to study from wherever, anytime, when

Motivation-to-E-Learn: A Quantitative Design Technique

Table 1. Means and standard deviations

Response Scale: (1) Of no importance to me; (2) Of very little importance to me: (3) Of little importance to me; (4) Of moderate importance to me; (5) Of importance to me; (6) Of high importance to me

I want it to,” “Using own computer when e-learning,” “Usefulness of this e-learning experience regarding my learning objectives,” “e-learning contribution to my competency development,” “Own experience with e-learning,” and “Others’ experience with e-learning.” Table 1 shows means and standard deviations calculated for the remaining fifteen variables. The highest agreement among students regarding their important motivation-to-e-learn variables happened on “Instructor support” and also on their personal belief about the adequacy of the communication with instructors when learning online. Next, data were processed by using maximum likelihood method of estimation for factor loadings. This is a frequently used method to search for hypothetical common factors that also gives significance test for the extraction of each successive factor (Nunnally, 1978). To simplify interpretation of the resulting factor structure, we apply varimax rotation and used only factor loadings greater than 0.30. In this manner we identified a two-factor structure. We used two criteria to select the number of factors: (a) significant p-value, and (b) eigenvalue greater than one. Table 2 shows the resulting two-factor structure.

Moreover, reliability coefficients for both factors were calculated. They are around 0.78, which is above what is accepted for this kind of study (Nunnally, 1978). This two-factor structure explains about 96% of the variance. The first factor is a linear combination of seven variables whose loadings range from 0,26 to 0,69. It accounts for 81% of the common variance. This factor groups “Accessibility to contents,” “Security and data protection,” “Personalised feedback,” “Easy-to-use interface,” “Flexibility in content presentation,” “Aesthetic content presentation,” and “Belief: I can learn this subject online.” Thus, it is named “Courseware” of the motivation-to-e-learn construct. It is worth noting that these seven variables address system, people, and content-related issues. They suggest important traits to facilitate content “consumption.” The second factor is a linear combination of eight variables whose loadings range from 0,43 to 0,59. These account for an additional 16% of the common variance. This factor groups “Resource availability,” “Instructor support,” “Institutional support,” “Feeling part of learning group,” “Liking studying subject matter,” “Belief: E-learning



Motivation-to-E-Learn: A Quantitative Design Technique

Table 2. Resulting factor structure

contribution to my learning objectives,” “Belief: Adequate communication with instructors,” and “Author credibility.” These variables tend to reflect the institution’s readiness to support technology-supported initiatives, in terms of IT investment, design choices, usability goals, and availability of resources. This readiness influences individual perceptions through the dissemination of related information among potential audience within its boundaries. That is why this factor is named “Organizational communication.” In practice, this should be taken as the preparation of learning context and students before deploying technology-supported learning initiatives. Three aspects are worth noting in this two-factor structure. First, seventy-three percent of the loadings are greater than 0,5 which indicates the presence of “pure”’ variables within the common factor structure. By extension, this strongly suggests content validity (Nunnally, 1978). Second, “Courseware factor” explained four times more variance than the other factor. This means that “Courseware factor” is of high importance to students and could strongly affect their motivation-to-e-learn. This is consistent with the technology acceptance model and its findings, which



empirically support the relationship between the technology’s usability and intentions to behave (Vantatesh et al., 2003). Last, the second factor suggests that a situated courseware could address some student concerns when learning online. Indeed, ensuring adequate conditions for strategic and operational technology-supported learning within contexts is an important requisite for an adequate people and courseware fit (Walker, 1992; Dix et al., 1998; Preece et al., 2002). This two-factor structure gives indication to prioritise action takings. To this end, designers could focus on the variables that heavily load on first factor, because they have a stronger impact on motivationto-e-learn than the others’ loading on the second factor. This does not mean that organizational communication variables can not be improved. At this matter, decisions must consider short- and long-term feasibility and resource availability at the initial deployment. After that, the remaining items can be used as a target for monitoring and improvement. Figure 3 shows a graphical representation of this structure as mapped into the theoretical framework of Figure 1. To know if prioritised action takings make a difference for students, designers, and instruc-

Motivation-to-E-Learn: A Quantitative Design Technique

Figure 3. Exploratory two-factor structure and conceptual framework

Student

Supporting system

Interaction

. Courseware . Organizational communication

Figure 4. Factor scores of motivation-to-e-learn factors

tors needs further information. By calculating factor scores for each participating student, they can identify specific intervention strategies to deal with contextual, courseware, or individual aspects that may affect learning results of existing student groups. For instance, as can be seen in Figure 4, enhancing “Instructor support” could affect students that scored positively on the organizational-communication factor. However, intensifying effort on giving feedback, making more flexible and aesthetic learning contents, and making easier-to-use interfaces could have a strong impact for many more students. Later, this analysis could be complemented by system usage data and other learning performance indicators. In this way educational policies and strategies, supported by technology, could be devised to improve the adequacy between learning-design

process, supporting system, student skills and concerns. Figure 4 shows factor scores for motivation-to-e-learn factors.

conclUsion anD FUTUrE WorK As previously mentioned, the goal of this work was to identify underlying dimensions of the motivation-to-e-learn construct. Our results show a two-factor structure: courseware and organizational communication. These factors represent a set of unmeasured (or hypothetical) underlying dimensions. Such factors are influenced by the variables that load on them. For instance, courseware factor is influenced by the fluctuations of the variables that heavily load on it, namely “Accessibility to contents,” “Security



Motivation-to-E-Learn: A Quantitative Design Technique

and data protection,” “Personalised feedback,” “Easy-to-use interface,” “Flexibility in content presentation,” “Aesthetic content presentation,” and “Belief: I can learn this subject online.” Since motivation-to-e-learn may be affected by events taking place in the surrounding environment during the learning experience, this two-factor structure model can derive direct implications on learning results. This might help set up a framework for focus on continuous improvement on what students’ value the most. Indeed, the identified two-factor structure suggests that the management of instructional contexts should focus on articulating courseware development and organizational communication to properly influence, through several channels, student perceptions on e-learning. Towards this end, intervention strategies should follow for each identified student groups. In this manner, all students could be “on board” with a systematic and cost-effective organizational effort. A major limitation of this study relates to the monocultural trait of the participating sample (Portuguese engineering students). Despite this limitation, our work comes as an attempt to quantitatively analyse underlying dimensions of motivation-to-e-learn. Despite that our findings are still preliminary, this factor structure already offers a quantitative way to holistically plan or improve technology-supported learning initiatives. To make this two-factor structure effectively useful as a design technique, future work must focus on validating it across contexts, learner groups, and learning situations to verify its generalizability. In addition, even though the reliability coefficient for each motivation-to-elearn factor is above 0.7, its stability over time should be also examined. Moreover, the usage of this scale to cluster student groups and support the definition of specific intervention strategies should be further explored.



rEFErEncEs Andersen, T. (2004). Toward a theory of online learning. In T. Andersen (Ed.), Theory and Practice of Online Learning. Athabasca University. Bandura, A. (1997). Self-efficacy: The exercise of self-control. New York: W.H. Freeman and Company. Dix, A., Finlay, F., Abowd, G., & Beale, R. (1998). Human computer interaction (2nd ed.). England: Prentice Hall Europe. Britain, S., & Liber, O. (2004). A framework for pedagogical evaluation of virtual learning environments (Technical Report). Bolton Institute for Higher Education. Costabile, M., De Marsico, M., Lanzilotti, R., Plantamura, V., & Roselli,T.(2005). On the usability evaluation of e-learning applications. In Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS’05) (Vol. 1, p. 6.2). Washington, DC: IEEE Computer Society. Gagne, M., & Deci, E. (2005). Self-determination theory and work motivation. Journal of Organizational Behaviour, 26, 331-362. Johnson, R. A., & Wichern, D. W. (1992). Applied multivariate statistical analysis (3rd ed.). New Jersey: Prentice Hall. Oppenheim, A.N. (2001). Questionnaire design (New ed.). New York: Irwin. Nunnally, J. (1978). Psychometric theory (2nd ed.). USA: McGraw-Hill Publishing Company. (Original work published 1967). O’Neill, K., Singh, G., & O’Donoghue, J. (2004). Implementing e-learning programmes for higher education: A review of the literature. Journal of Information Technology Education, 3, 312-323.

Motivation-to-E-Learn: A Quantitative Design Technique

Organ, O., & Bateman, T. (1991). Organizational behaviour (4th ed.). USA: Irwin. (Original work published 1978). Preece, J., Rogers, Y., & Sharp, H. (2002) Interaction design: Beyond human-computer interaction. New York: John Wiley & Sons. Rentroia-Bonito, M. A., Jorge, J.A., & Ghaoui, C. (2006). Motivation to e-learn within organizational settings: An exploratory factor structure. International Journal of Distance Education Technologies, 4(3), 24-35. Rentroia-Bonito, M. A., & Jorge., J. A. (2004). Motivation to e-learn within organizational settings: What is it and how could it be measured? In Proceedings of 15th IRMA International Conference, New Orleans, LA. IRMA. Retrieved December 2005 from http://citeseer.ist.psu. edu/700830.html

Ventatesh, V., Morris, M., Davis, G., & Davis, F. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3). Walker, J. (1992). Human resources strategy. McGraw-Hill Series in Management. USA: McGraw-Hill. (Original work published 1980). Walker, S. (2003). Distance education learning environments research: A short history of a new direction in psychosocial learning environments (Technical Report). TCC. Wentling, T., Waight, C., La Fleur, J., Wang, C., & Kanfer, A. (2000). E-learning: A review of literature (Technical Report). University of Illinois (Urbana-Champaign).





Chapter V

Algorithm Education Using Structured Hypermedia Tomasz Müldner Acadia University, Canada Elhadi Shakshuki Acadia University, Canada Andreas Kerren Växjö University, Sweden

absTracT Understanding of algorithms is one of the most challenging aspects of the study of computer science. Over two decades of research has been devoted to improving techniques to learn and teach algorithms. In this work, we present a new approach for explaining algorithms that aims to overcome various pedagogical limitations of the current visualization systems. The main idea is that, at any given time, a learner is able to focus on a single problem. This problem can be explained, studied, understood, and tested before the learner moves on to study another problem. The structured hypermedia algorithm explanation (SHALEX) system is the system we designed and implemented to explain algorithms at various levels of abstraction. In this system, each abstraction is focused on a single operation from the algorithm using various media, including text and an associated visualization. The explanations are designed to help the user to understand basic properties of the operation represented by this abstraction, for example its invariants. SHALEX allows the user to traverse the graph-based algorithm model, using a top-down (from primitive operations to general operations) approach, a bottom-up approach, or a mix of these two approaches. Since the system is implemented using a client-server architecture, it can be used both through distance education and in the classroom setting. To aid and monitor the leaner, we also developed an agent in SHALEX that provides help and monitors the completion rate. Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Algorithm Education Using Structured Hypermedia

inTroDUcTion In the ninth century, Mukhammad ibn Musa AlKhorezmi, the chief mathematician in the academy of sciences in Baghdad, helped establish the Indian numbering system, using decimal notation and the zero. (Al-Khorezmi apparently came from the oasis of Khorazem, at the southern end of the Aral Sea, in what is now Uzbekistan.) When his work reached Europe, the Europeans misspelled the author’s name and called its use “algorism” or “algorithm.” Nowadays, an algorithm means a specific set of instructions for carrying out a procedure or solving a problem, and algorithms play an intrinsic role in computer science. Because of their importance, researchers have been trying to find the best way to teach and learn algorithms. One of the best known approaches has been to use visualization; especially its understanding as “the power or process of forming a mental image of vision of something not actually present to the sight” (Petre et al., 1998a). The following describes a typical approach used for algorithm visualization: 1.

2. 3. 4.

Take the description of the algorithm (usually this description comes in the form of the code in a specific programming language, such as C). Represent data graphically in the code using bars, points, and so forth. Represent the flow of control using animation. Illustrate the animated algorithm and hope that the learner will understand the algorithm.

Step 2 is often automatically generated from the source code, possibly with the help of specifying “interesting events” (Brown & Sedgewick, 1998).

Although, there have been several attempts by many researchers describing the use of animation to software explanation, we will not review these usages. Interested readers are referred to the following two best-known anthologies: Gloor (1992) and Gloor (1998). There are various problems with the above approach. First, providing a graphical representation of an algorithm is just another way to show the code of the algorithm—instead of using a textual programming language, we use a graphical language. Executing the visualization, we simulate the code written in a textual language, using a graphical language and the representation is typically at the low level of abstraction that shows the low level steps. The work presented herein describes an alternative and systematic procedure to explain algorithms that can be used in various settings, including continuing education. Our approach is based on results of evaluations of existing algorithm visualizations, some findings from cognitive psychology and constructivism theory regarding the active use of algorithm explanation systems, and experience from software engineering and verification regarding the use of multiple levels of abstraction and properties, such as invariants, to explain algorithms and justify their correctness. Based on the experience with design patterns (Gamma et al., 1995), we believe that algorithm explanation should not necessarily be prepared by experts, who have intimate knowledge of the algorithm and the best way of coding it. In this chapter, we describe a system called structured hypermedia algorithm explanation (SHALEX). This system provides several novel features, such as reflection of the high-level structure of an algorithm rather than low-level steps, and support for programming the algorithm in any procedural programming language. We use a graph-based algorithm abstraction model, in which algorithms may be studied top-down, bottom-up, or using a mix of the two. In this model,



Algorithm Education Using Structured Hypermedia

a single abstraction is designed to explain one operation using the abstract data type (ADT) defined for this specific abstraction, with the top abstraction explaining the algorithm. For example, the root abstraction for a selection sort algorithm is designed to explain this algorithm, using the ADT operations. In this example, one of the ADT operations is a function smallest() that finds the smallest element in the sequence. A visual representation is used by the student to help him or her understand the basic properties of this abstraction, for example, invariants of the selection sort. The lower levels of abstraction focus on operations that are considered primitive at the higher levels. In the selection sort example, the low level abstraction provides ADT with primitive operations to represent the function smallest(). Accordingly, some operations are implemented at lower levels of abstraction, while others are left as primitives. While the code is presented as text, there is also an associated visual representation, which is used by the student to help him or her understand the basic properties of this abstraction. Complex animations are not used because experimental data show that animations may confuse students (Stasko & Lawrence, 1998). In addition, SHALEX includes a learner model to provide spatial and temporal links, and to support evaluations and adaptations. The learner using SHALEX would study algorithms by interacting with a system for algorithm explanation. An algorithm explanation system is an interactive system, designed to include a student model to support evaluations and adaptations. It guides the student through levels of abstraction, and asks related questions; for example what are the invariants of the algorithm? Note that the design and implementation of the system makes it suitable for both traditional classroom/lab setting and for distance education, where teacher/student interactions are limited.

0

rElaTED WorK An algorithm animation (AA) visualizes the current state of the algorithm and animates transitions between successive states [whereas algorithm visualization (AV) also covers static visualizations, such as flowcharts]. There is an essential difference between AA and program animation (PA), in that the former visualizes algorithms while the latter visualizes the actual program execution written in a concrete programming language. They both are areas of study known together as software visualization (SV); see Price et al. (1993). There has been a lot of research on algorithm animation, see the introduction chapter of Kerren and Stasko (2002) in the book (Diehl, 2002), and also see Eades and Zhang (1996) and Stasko et al. (1998). Other aspects of algorithm animation, including taxonomies are described in Brown (1988), while the use of abstraction is covered in Cox and Roman (1992) and the user interface issues are tackled in Gloor (1998b). Figure 1 illustrates a typical example of an algorithm animation—a screenshot from the animation of insertion sort using JAWAA, taken from Whitley & King (2001). The animation is accompanied by the following instructions: •

•

•

Start by comparing the first two rectangles in the group, and shift the second rectangle to the left if it is smaller. Look at the next rectangle, compare with the one on its left, and shift it to the left until it is smaller than the one to its right and larger than the one to its left. Repeat step 2 until finished.

It is clear from Figure 1 that the user has to map the problem domain (values to be sorted) to the graphical domain (bars), and then looking at the animation the user has to retrieve essential properties of the algorithm, such as maintaining a sorted prefix. Such algorithm animation systems resemble visual debuggers in that they show the

Algorithm Education Using Structured Hypermedia

Figure 1. Animation of insertion sort

execution of the algorithm by code-stepping, work at the lowest level of abstraction, and illustrate only the primitive code. This approach constrains users to view the code in the order of execution, which is the wrong information for understanding the algorithm. In addition, it has a poor cognitive fit with the plan-and-goal structures that users are trying to extract from the code (Petre et al., 1998a). Finally, these systems suffer from the lack of focus on relevant data (Braune & Wilhelm, 2000). Besides JAWAA, there are many similar systems used for visualization. JSamba (1998) is the Java version of the well-known SAMBA (Stasko, 1996) AV system that uses a scripting language to support graphic primitives and animations. On the other hand, interactive data structure visualization (IDSV) proposed in Jarc et al. (1995) is not a general-purpose system; instead it provides animations of various algorithms, for example sorting algorithms. IDSV is more interactive than JAWAA and JSamba. It also provides “I’ll try” function to test users’ understanding of the

algorithm. The tests are performed by asking users questions such as “What is the next step of the algorithm?” Another system called “a new interactive modeler for animations in lectures” (ANIMAL) proposed by Rößling et al. (2000) is a general purpose AV system. In this system, animation can be created by scripting, calling Java-based API, or using a specialized editor. The JHAVÉ system (Naps, 2005) is a support environment for a variety of available AA systems, which provides several interaction support tools, such as input generators, stop-and-think questions, VCR controls, and so forth. The algorithms in action (AIA) system proposed in Stern and Naish (2002a; 2002b) does not use an abstract, language-independent pseudocode and has some features that belong to a visual debugger rather than to a tool to learn algorithms. Note that it is also possible to embed any Web-viewable animations built by AA systems, for example see Ganimal (Ganimal, 2007; Diehl et al., 2002; Diehl & Kerren, 2002), Animal (Rößling & Freisleben, 2002), or JSamba



Algorithm Education Using Structured Hypermedia

(JSamba, 2007), as well as other formats, such as ones used in Macromedia Flash (Macromedia, 2007) that include visualizations, animated GIFs, sound files, and so forth. Finally, Baloian et al. (2005) suggest using so-called concept keyboards (CKs). Each key of a CK is mapped to the execution of an existing method available in the implementation of the input algorithm. The available keys may be different at various stages of the execution, so that the user could trigger operations at the required level of abstraction. The authors performed several software evaluations suggesting that the active use of CKs leads to a better understanding of how algorithms work. Evaluations of systems designed to explain algorithms using various visualization and animation techniques have not shown that these systems are always educationally effective (Hundhausen et al., 2002). Indeed, some studies found that the effect of using animation is either neutral or even negative (Stasko & Lawrence, 1998). Dijkstra (1989) even feared “permanent mental damage for most students exposed to program visualization software.” However, Crosby and Stelovsky (1995) found that students who interacted with an algorithm animation performed significantly better than students who listened to a lecture. It should be noted that software evaluations are difficult to verify, and widely used test designs have various disadvantages, see Baumgartner (1999). Therefore, various researchers tried to identify educational problems, particularly evident in continuous education settings, where the learner cannot always count on immediate help of the instructor. In what follows, we describe these attempts. First, Biermann and Cole (1999) determined that multiple views showing algorithm states are helpful to avoid forcing the viewer to remember the previous states. The second attempt centers on the use of abstraction. Algorithms represent abstract processes but this aspect is rarely considered. One approach



presented by Wilhelm et al. (2002) uses a static source code analysis to abstractly execute the algorithm on “all possible sets of input data,” and visualize invariants. An extension of this approach was exemplified for binary tree algorithms; see Johannes et al. (2005). A related idea is to use multiple levels of abstraction; and this idea is supported by Petre et al. (1998b) who claim that in general it is hard to determine a single suitable level of abstraction. Their research has shown that if the presentation is designed to highlight some kind of information, then it is likely to obscure other kinds. The third issue is the abstract model of the algorithm. Here, one attractive idea is to use pseudocode, with which the algorithm can be studied independently of any programming language (see Fleischer & Kucera, 2002; Naps, 2005). The pseudocode may have an additional visual representation, which exposes its properties, in particular, its invariants. Finally, pseudocode may be designed at various levels of abstract data structures and operations (designed so that they can be directly mapped to most procedural and object-oriented programming languages). The fourth attempt at improving the current AV systems is based on the constructivist idea that the knowledge has to generate itself in the learner’s mind, and so it cannot be transferred in a traditional way, for example, by instruction. Constructivism principles are based on active learning (Hundhausen et al., 2002) and this style of learning includes various kinds of interactions with the learner. For example, students are able to use their own input data sets, use a do-it-yourself mode and predict the next step of the algorithm, or determine the essential algorithm properties. The moderate constructivism proposes a system where the teacher, the expert and the system are not allowed to manipulate the learner’s construction process but they can offer help and coach their individual construction processes. Algorithm explanations should not be prepared

Algorithm Education Using Structured Hypermedia

by experts; instead they should be prepared by learners themselves. This approach was used in some explanation systems, for example in studying compiler generation techniques to generate interactive algorithm animations from specifications (see Kerren, 2004a,b). Visualizations use various kinds of multimedia, including graphics to represent data, animation and video to convey the temporal evolution of a computer algorithm (Stasko & Lawrence, 1998), and voice for explanations, also called auralization (Brown & Hershberger, 1998). Since an algorithm is a process that is both abstract and dynamic, the system designed to explain algorithms should emulate both these features. Therefore, the fifth, and relative new, research area that attempts to produce more effective algorithm explanation systems suggest using hypermedia. For example, HalVis showed the advantage of using hypermedia over using just animations (Hansen et al., 2002). In particular, this system provides hyperlinks that help the learner to move between various kinds of descriptions, for example, text and animations; and finally the analogical animation, including both micro and macro-animations. However, HalVis allows the users to learn in one direction using a top-down approach, which does not always reflect the structure of the algorithm and is not adaptive. Additionally, it supports abstractions, but only for micro/macro-level animations. The aforementioned algorithm animation system, called Ganimal (Diehl & Kerren, 2002; Ganimal, 2007) supports hypermedia, through the use of a specialized algorithm animation specification language, called Ganila. This language offers a set of control structures, which for example can be used to annotate the statements of the underlying algorithm with URLs. Ganila programs are translated into Java and executed within its own runtime system for animation, and showing visualizations available at specified URLs. In summary, most existing systems do not attempt to visualize or even suggest essential properties, such as invariants, which are essential

for understanding algorithm correctness. One notable exception is the approach in Wilhelm et al. (2001), which uses a static source code analysis to abstractly execute the algorithm on “all possible sets of input data,” and visualize invariants.

algoriThm EXPlanaTion This section describes in details our proposed approach, which is called algorithm explanation (AE). The main goal of AE is to support the task of algorithm comprehension (we are interested in learning, rather than teaching), see Müldner and Shakshuki (2006). According to Petre et al. (1998b), to make this task possible, the learner has to build a mapping from the domain consisting of the algorithm entities and temporal events to learner’s conceptions of these entities and events. Our target audiences are students who know well programming in at least one programming language, such as C, C++ or Java and are willing to learn algorithms. These students need to learn our conventions that are used for writing the pseudocode and showing visualizations. Design Goals In this section we describe what exactly we want to achieve by using AE, that is, what we expect the student will learn. This can be described by the following goals: G1: Understanding of both what the algorithm is doing and how it works. G2: Ability to justify the algorithm correctness (why the algorithm works). G3: Ability to program the algorithm in any programming language. G4: Understanding of time complexity of the algorithm. In order to describe a system that satisfies these goals, we now list the requirements that an algorithm explanation should satisfy.



Algorithm Education Using Structured Hypermedia

Design Requirements To achieve the above goals, the following requirements must be satisfied: R1: The algorithm is presented at several levels of abstraction. R2: Each level of abstraction is represented by the pseudocode, and optionally by visualizations. R3: Active learning is supported. R4: The design helps to understand time complexity. R5: The presentation uses multiple views. R6: Presentations are designed by experts. Now, we elaborate on each of the above requirements and explain how they are related to our goals. Re R1. There are several advantages of using multiple levels of abstraction. First, the research in cognitive psychology on knowledge organization supports using multiple levels of abstraction when dealing with complex tasks (Anderson, 1980). The idea of using more than one level of abstraction is also supported by Petre et al. (1998b), who claim that in general it is hard to determine a single suitable level of abstraction. Second, the research has shown (Petre et al., 1998b) that if the presentation is designed to highlight some kind of information, then it is likely to obscure other kinds. In our approach, each level of abstraction is used to highlight a single kind of information; for example a single invariant, and so the student can focus on this kind of information. Third, to reason about a process in the world requires setting up a mental model of each state of an algorithm (Petre et al., 1998b). There are two possible approaches to defining levels of algorithm abstraction. For both approaches, the top level is first defined, using operations considered to be primitive at this level. Then, with the first approach, each top level primitive is defined at the single lower abstraction level,



possibly using other primitives (those primitives would be defined at the next lower abstraction level). With the other approach, the top-level primitives are replaced (rather than defined) by their definitions, which again can use some lower-level primitives. Thus, the second approach resembles code inlining. We have chosen to take the former approach, which helps to concentrate at issues pertinent to one abstraction level. We disagree with the criticism in Faltin (2001), which claims that structuring algorithms using this approach produces code that is longer and less efficient. The recent advances in compiler technology and the right mapping of the correctly designed pseudocode to a programming language make the code sufficiently efficient (see the illustrative example section). The latter approach was investigated in Feng (2003). It should be noted that, in our approach, each level of abstraction has a small number of states. This is important because the large number of states makes it difficult to reason about properties of the algorithm. There are three possible ways to studying an algorithm: top-down, bottom- up and a mix of two. We advocate top-down approach, starting from the top level of abstraction and then moving down to lower levels. The reason for using a top-down approach is that it helps students to understand various special cases, such as various ways to compare integer values in a sorting algorithm, whereas a bottom-up approach hardcodes specific cases. This approach differs from the one used in “algorithms in action,” AIA, from Stern and Naish (2002a, 2002b), where students were allowed to choose the starting level of abstraction, and often chose a bottom-up approach. Therefore, this requirement contributes to G1, in both “what” and “how.” Re R2. The pseudocode is a model of the algorithm, and it includes the high-level abstract data structures and operations. These operations are designed so that they can be directly mapped to most procedural and object-oriented

Algorithm Education Using Structured Hypermedia

programming languages. Using pseudocode, the algorithm can be studied independently of any programming language (Fleischer & Kucera, 2001). The pseudocode given at each level of abstraction has an optional visual representation, which exposes its properties, in particular, its invariants. This approach is supported by findings provided by Petre et al. (1998b) that proper explanation supports a display-based reasoning; that is, the display becomes a focus for reasoning and supports creating the mental image of things that do not really appear there. For example, the visualization in Figure 8 helps to recognize two invariants of this algorithm. Note that our concept of a pseudocode differs from that used in Stern and Naish (2002a, 2002b), where the pseudocode is based on C, with some abstract procedure calls. By exposing the algorithm’s properties, in particular its invariants, at various levels of abstraction, this requirement contributes to G1 and G2. In addition, the pseudocode is written in generic terms so that it can be used not only to write concrete implementations in specific programming languages such as Java, but also to produce different concrete implementations; for example, using linear or binary search to implement insertion sort. This contributes to G3. Re R3. Active learning follows cognitive constructivism principles (see Hundhausen et al., 2002). This style of learning includes various kinds of interactions with the student. For example, students are able to use their own input data sets; use a do-it-yourself mode--that is, predict the next of the algorithm (Faltin, 2001; Stern, 2002a; Stern, 2002b), and determine the essential algorithm properties. AE always comes with several sets of representative sample input data; the learners can also use their own input data. AE includes student evaluation, which consists of programming the target algorithm, a textual programming language. This ability is missing from all algorithm animation systems, but in our opinion it is absolutely essential. The

ultimate goal of teaching algorithms is to educate programmers, who will be able to implement various algorithms. Finally, AE includes post-test (Stasko & Lawrence, 1998) with tasks such as hand-execution of the algorithm on sample sets of input data, and answering various questions about the algorithm. Re R4. AE helps the learner by providing tools that help to understand time complexity, described in the next section. Re R5. Multiple views showing algorithm states are used to avoid forcing the viewer to remember the previous states [two consecutive frames are shown in the “comic strip” approach (Biermann & Cole, 1999)]. The decreased cognitive overhead contributes to understanding how the algorithm works, which contributes to goal G1. Re R6. AE presentations are designed by experts who have a complete understanding of essential properties to be exposed, such as invariants. Essentially, our approach is similar to that used in design patterns, which are created by experts and used by novices. It is also similar to the concept of “algorithmic design patterns” described in Goodrich and Tamassia (2001). The design by experts leads to a better code. Note that our proposal differs from that proposed by Hundhausen and his co-workers (Hundhausen et al., 2002), who recommend that students should design algorithm animations. We strongly believe that designing algorithm animations concentrates on meta-tools and distracts students from their primary goal, which is learning algorithms. AE uses a variety of tools to help the students to learn algorithms, and visualization is just one possible tool. Indeed, we provide both textual and visual representation and use the latter to help students recognize and understand algorithm properties. Note that Petre et al. (1998b) stated that two representations are not necessarily better than one. In this work, we use a visual representa-



Algorithm Education Using Structured Hypermedia

tion to help students derive essential properties, such as invariants. However, we leave the option of using just the text, and if the students can successfully derive invariants from the text they do not have to see the visual representation. Thus, the visual representation is used to help reason about the textual representation. It is available to the user, because in some cases this representation provides the so-called “gestalt” effect; it provides an overview making a structure clearer (Petre et al., 1998b). Algorithm Explanation Descriptions One of the main goals is to develop an AE catalog consisting of descriptions of many well-known algorithms. Each description, or catalog entry, consists of the following four parts: 1.



A hierarchical abstract algorithm model (AAM) consists of abstractions representing operations. Each abstraction explains a single operation op(), and consists of a textual representation and an optional visual representation. The textual representation includes an ADT that provides data types and operations. It also provides a representation of the operation op() using the ADT from this abstraction. All these abstractions form a hierarchical graph (in many cases a tree) that has a single dedicated root node representing the abstraction of the algorithm operation. The AAM graph is used only by the implementation of the explanation system and its complexity is transparent to the learner, who is guided by the AE system (see Design of AE Systems subsection). There are three possible modes that can be considered for explaining an algorithm. In the first mode, the explanation can start from the algorithm explanation and proceed towards more primitive operations; this mode is represented by the top-down traversal of the AAM. In the second, the explanation can start from primitive operations and

2.

3. 4.

proceed towards the algorithm explanation; this mode is represented by the bottom-up traversal of the AAM. The third mode, called learner-selected mode, is a mixture of the both modes. For any operation op that appears in the operation currently focused on, the learner may select op and request one of the following: help, taking a test (if the author decided to include testing), or explanation of this operation. In the first case, SHALEX provides a context-sensitive help. In the second case, the learner may be given a test, and if the test is passed, the learner model will be updated. The author may specify that in order to complete studying the algorithm, the learner has to complete all tests, using evaluations available in the author model. An intermediate representation of the socalled primitive operations from the ADTs, which are not implemented in the AAM. This representation is designed to help the student to write concrete implementations. Tools that can be used to help predicting the algorithm complexity. Questions for students, including “do it yourself” mode.

To explain part 1, we assume that f() is an operation. The abstraction that explains f(), abst(f) is a pair (ADT, representation of f() in the ADT), where ADT consists of data types and primitive operations, see Figure 2. An abstraction abst(f) is a parent of an abstraction abst(g) provided that g is one of the primitive operations from the ADT abst(f). Therefore, a child abstraction provides a partial implementation of the operation from the parent abstraction. Typically, there are only few operations from any abstraction’s ADT that are implemented in a child of this abstraction; others are considered primitive operations. An abstract algorithm model (AAM) of an algorithm f() is a hierarchical graph rooted at abst(f). For example, the AAM of a selection sort is a tree of

Algorithm Education Using Structured Hypermedia

Figure 2. Abstraction node of the AAM AADDTT abst(f) abst(f) f

•• DDaata ta Typ Typeess •• O Oppeera ration tionss

•• AADDTT •• RReeppre rese senntation tationss

RReeppre rese senntation tationss •• •• ••

abstractions rooted at abst(selection). To explain an algorithm, we construct an AAM tree with sufficient number of levels so that the student is able to understand how and why the algorithm works. In particular, the student can form and justify invariants of the algorithm. The intermediate representation of all AAM’s primitive operations is provided in Part 2. To implement the algorithm in a specific programming language, the student has to map all primitive operations that do not have implementations in the AAM to the selected language. This representation is called an abstract implementation model (AIM). The representations in AIM are generic in that they are not using any specific programming language; instead they use high-level concepts that can be mapped to many languages. Once all primitive operations are represented in the AIM, the code for the algorithm and possibly some of its operations can be easily represented using this AIM. Let’s add that more than one explanation of a single algorithm can be made available in our system, for example the AAM graph for Insertion Sort shown in Figure 3 can be expanded by adding explanations from the INSERT ADT. Then, the learner can choose a version with fewer or greater number of explanations. Instead of the learner making this decision, it is desirable to have a system that reacts to the learner’s progress. Part 3 deals with an explanation of algorithm

Te Text xt VVisu isualiza alizatio tions ns Im Im pple lem m en enta tatio tions ns (A (AIM IM))

complexity, which is one of the most difficult goals of algorithm visualization. This is because it requires mathematical notions that are hard to visualize. One of the rare approaches in this direction is described in Pape and Schmitt (1997). There may be three kinds of tools designed to help the student to derive the complexity of the algorithm being studied. The first tool, based on Horstmann (2001), gives the student a chance to experiment with various data sizes and plot a function that approximates the time spent on execution with these data. The second tool, based on Goodrich and Tamassia (2001, p. 477), provides a visualization that helps to carry out time analysis of the algorithm. The third tool, also related to part 4 above, asks students various questions regarding the time complexity, and questions specific to the algorithm being studied, and evaluates their answers (for the example, see the Post Test subsection). The current version of SHALEX supports these tools and different approaches. Graph Representation of AAM To explain an algorithm, it is often necessary to explain various abstract data types (ADTs). As an example, consider a Kruskal algorithm to find a minimum-cost spanning tree (Aho, 1983). This algorithm uses three ADTs: SET, PRIORITY-QUEUE and MERGE-FIND. Each of these ADTs includes a number of operations that have to be explained. Note that there are two kinds of



Algorithm Education Using Structured Hypermedia

Figure 3. The AAM for Insertion Sort root n ode

aabbsst(in t(inse sertio rtionn))

AAD DTT

aabbsst(in t(inse sert) rt)

R Reeppre rese senntation tationss

AADDTT R Reeppre rese senntation tationss

aabbsst(sw t(swaapp)) prim itive op eration

Figure 4. Graph representation of the Kruskal algorithm aabbsst(kruskal) t(kruskal)

PPRRIO IORRIT ITYY-Q QUUEEUUEE

SSEETT

M M EERRG GEE-FFIN INDD

LLin inke kedd-Lists -Lists

Tre Treeess

aabbsst(m t(m eerge rge))

ADTs: “pure” ADTs that define operations but no implementations of these operations, and “concrete” ADTs that define specific implementations. For example, a pure ADT PRIORITY-QUEUE can be implemented as a concrete ADT that uses singly linked lists. Using Java terminology, a pure ADT is an interface, whereas a concrete ADT is the implementation of this interface. To be able to reuse ADTs, we introduce two kinds of nodes in the AAM, namely a type node and an operation node. The explanation provided by the AAM graph shown in Figure 4 gives the



aabbsst(fin t(findd))

textual and visual representation of the Kruskal algorithm using three pure ADTs, shown in uppercase. Each of these ADTs may have different implementations, shown in lowercase. For example, PRIORITY-QUEUE can be implemented using a linked list, and MERGE-FIND can be implemented using trees. Note that in Figure 4 there are five type nodes (shown in boxes with single frames), and three operations nodes (shown in boxes with double frames). The “Linked List” type node has two parents, indicating that it can be used to implement ADTs from both of these parents.

Algorithm Education Using Structured Hypermedia

algorithm Explanation system Design It is expected that an AE system should help the student to select one algorithm. It should also guide the student to go through its explanations. These explanations are stored in the AAM, and the student will follow these steps if we assume that top-down learning was selected (recall that the actual representation of AAM is transparent to the student who is guided by the AE system): 1.

The root abstraction is explained: a. The ADT associated with the root abstraction is shown. b. The implementation of the operation from this abstraction, using the ADT operations is shown. c. The student is asked to explain basic properties of the implementation from (b), including invariants. d. The student may choose to see the visualization associated with the

2.

3.

implementation from (b); if so this visualization is made available, and the student can enter specific input data and watch the visualization, or use a “do it yourself” mode to test their understanding of the algorithm. e. Explanations as to which ADT operations are primitive operations. For any child abstraction, associated with the higher-level operations that are not primitive, the explanation process from 1 above is repeated. Now, the student is supposed to implement an algorithm in a selected programming language. First, AIM is shown, and the student is asked to implement an algorithm in one of several available programming languages. Sample input data are provided, including boundary cases that can be used by the student for testing and debugging their implementations.

SHALEX is a realization of this architecture. At first, it offers students a window where they

Figure 5. Choosing an algorithm and ADTs from the AAM



Algorithm Education Using Structured Hypermedia

Figure 6. Learning the specific child operation insert

can select an algorithm from a list (in our example only one selectable). Then, a second window with the current AAM appears, compare Figure 5, and the student can follow steps described before from the root node to different child nodes, and then to the leaf nodes. Figure 6 shows this learning situation [steps (a)-(e)] of the child node “insert.” By clicking the “View” button at the bottom of the window, the learner can study an associated visualization as described in step 1 (d). Algorithm visualization systems can generally be categorized using a taxonomy introduced by Price et al. (1998). Using this taxonomy, the above system is a specific system (used only for limited number of algorithms), and its content, which defines what aspect of an algorithm is visualized, concentrated on explaining algorithm properties. The form, which is how a system is presented in text, graphics, and animation, and the method, which describes how to develop an explanation, are currently hand-coded from scratch. Finally, the interactions include describing invariants, answering specific questions, and implementing the algorithm. 0

illUsTraTiVE EXamPlE oF algoriThm EXPlanaTion DEscriPTion To demonstrate and explain our proposed approach, an example of algorithm explanation using selection sort is provided in this section. Selection Sort Example Recall that an algorithm is explained using various levels of abstraction including root and leaf levels of abstractions; each abstraction is designed to present a single operation used in the algorithm. This process is explained using the selection sort (Aho et al., 1983). Root Abstraction The root abstraction provides the ADT and the representation of the algorithm using this ADT. Abstract Data Type The ADT consists of data and operations. The data consists of sequences of elements of type T, denoted by Seq, with a linear order. In general, this order can be defined in one of three ways:

Algorithm Education Using Structured Hypermedia

1.

2.

3.

Type T supports the function int compare(const T x) so that y.compare(x) returns -1 if x is less than y, 0 if they are equal and +1 otherwise. Type T supports the “ 0.70). The results were then used for a factor analysis using a varimax rotation. A value of λ > 0.40 was assumed to show strength of relationship between the independent variables (survey questions) and the dependent variables (constructs).

Perceived Usefulness 48% of the sample agreed that online learning will be beneficial for them to get prepared for a check ride, whereas 28% were neutral about it and 24% responded that it will be not be beneficial to take online training. The study found that 62% of the sample agreed that online learning improves their ability to get prepared for a check ride, whereas 16% were neutral about it and 22% did not agree. 54% of the sample agreed that the advantages of online learning outweigh the disadvantages, whereas 26% were neutral and 20% were negative. However, as illustrated in Figure 3, the response is strongly positive towards online learning being advantageous. Moreover, 66% of the sample agreed that, overall, online learning presents advantages for them, whereas 16% were neutral about it and 18% thought it would not be advantageous for them. Analysis of the perceived usefulness (PU) resulted in an unacceptable alpha of 0.6343. However, anecdotal evidence from pilot interviews shows that pilots are highly receptive to the possibility of online learning as a substitute for current traditional methods, but could quickly turn if the product does not provide interactivity, flexibility, and useful knowledge.



Distance Learning in Business Aviation Industry

Figure 3. Histograms for the perceived usefulness construct Q

Q



0

  0 

0

 Std. Dev = .0 

Mean = .

0

N = 0.00 .0

.0

.0

.0

.0

.0

Frequency

Frequency



Std. Dev = 1. Mean = 4. N = 0.00

0

.0

.0

Q

.0

.0

.0

.0

.0

.0

Q

Q

Q



0



0

 0 

Std. Dev = .



Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

Frequency

Frequency

 Std. Dev = . Mean = . N = 0.00

0

.0

.0

Q

.0

.0

.0

.0

.0

.0

Q

Figure 4. Histograms for the compatibility construct Q 0

0

0

Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

compatibility 60% of the sample agreed that online learning fits well with the way they take ground training, whereas 14% were neutral about it and 26% thought it will not be compatible with the existing way they are taking the ground training. Subsequently, 70% of the sample agreed that online learning is compatible with their work



Frequency

Frequency

Q 0

Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

style, whereas 12% were neutral about it and 18% think it will not be fit with their existing work style. Figure 4 illustrates the histograms for the compatibility construct.

Ease of Use The study reveals that 58% of the sample responded that online learning would not be more

Distance Learning in Business Aviation Industry

difficult to follow than current training techniques, whereas 26% were neutral about it and 16% thought it would be difficult to follow. 44% of the sample responded that online learning would not require greater mental effort than the current training environments, 32% were neutral about it and 24% think it would require lot of mental effort. In addition, 36% of the sample responded that online learning was not frustrating, whereas 28% were neutral about it and 36% thought it is often frustrating. The authors assume that

the response regarding frustration was made in comparison with the existing way they are taking ground training. During pilot interviews a negative feedback towards online learning due to a perceived lack of interactive, audio, visual, and tactical features was noted. Computer-based training such as SimuFlite’s FasTrack has caused frustration and boredom for pilots in the past. 60% of the sample respondents expect online learning to be clear and easy to understand, 20% were neutral and 20% do not expect that it to be

Figure 5. Histograms for the ease of use construct Q

Q



0



0

 0 

Std. Dev = .



Mean = . N = 0.00

0 .0

.0

.0

.0

.0

Frequency

Frequency

 Std. Dev = . Mean = . N = 0.00

0

.0

.0

Q

.0

.0

.0

.0

.0

Q

Q

Q0











0

0   



 Std. Dev = . 

Mean = . N = 0.00

0 .0

.0

.0

.0

.0

Frequency

Frequency



Std. Dev = .



Mean = . N = 0.00

0

.0

.0

Q

.0

.0

.0

.0

.0

.0

Q0

Q Q



0

  0

0

 

Std. Dev = . Mean = . N = 0.00

0 .0

Q

.0

.0

.0

.0

.0

Frequency

Frequency

0

 Std. Dev = .0 

Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q



Distance Learning in Business Aviation Industry

Figure 6. Histograms for the peer influence construct Q

Q

0

0

0 0

0

Std. Dev = .0 Mean = . N = 0.00

0 .0

.0

.0

.0

.0

Q

clear and easily understandable. The positive response clearly indicates that the majority of respondents expect online learning to be user friendly and easy to operate. Neutral responses may indicate that they are not sure what it will be like, whereas negative feedback towards clarity and understandability of online learning may be due to their previous experience or lack of awareness of emerging online learning technologies. Subsequently, 78% of the sample respondents expect that online learning will be flexible to interact with, whereas only 8% were neutral about it and only 14% do not expect it be flexible. The positive response clearly indicates that almost 80% of the respondents expect online training to flexible and tailored to their needs. This shows a very strong desire for flexibility in any online learning environment. Finally, the survey reveals that 62% of the sample respondent expects that online learning will increase their skill to get prepared for a check ride, whereas only 20% were neutral and only 18% do not expect it be flexible. The positive response clearly indicates that majority of the respondent expects online training will be able to help them to get prepared for check ride at their own ease and convenience. Figure 5 illustrates the histograms for the ease of use construct.



Frequency

Frequency

0 0

Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

Peer Influence 30% of the sample respondent did not agree that their friends will have any influence on them to take online learning to get prepared for a check ride, whereas 60% were neutral and only 10% believe that their friends can influence them to a certain extent. This negative response indicates that the respondents do not think that they will be influenced by their friends to take online learning. It was also observed that 32% of the sample respondent did not agree that their fellow pilots will have any influence to take online learning to get prepared for check ride, whereas 48% were neutral and 20% believe that their co-pilots can influence them to a certain extent for online learning. The negative response clearly indicates that one third of the respondents do not think that their fellow pilots will have any influence on taking online learning, whereas the majority of the respondents provide neutral response, which indicates that they are not sure whether their fellow pilots will have any influence on them. Figure 6 illustrates the histograms for the peer influence construct.

Supervisor Influence The survey reflects that superior influence was an important construct and advises that more complete research be conducted in the future. 28%

Distance Learning in Business Aviation Industry

Figure 7. Histograms for the supervisor influence construct Q 0

0

0

Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

Frequency

Frequency

Q 0

Std. Dev = . Mean = . N = 0.00

0

.0

.0

Q

.0

.0

.0

.0

.0

.0

Q

Figure 8. Histograms for the efficacy construct Q

Q

0

   0

0



Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

of the respondents agreed that their supervisor expects them to take online learning, whereas a 38% majority were neutral about it and 34% did not believe that their supervisor will expect them to take online learning. The positive response clearly indicates that 28% of the respondents assume that their supervisor might expect them that they should take online learning, whereas the majority of the respondents provide neutral response, indicating that they are not sure whether their supervisor will have any say regarding the interest for online learning, and one third of them eliminated the possibility of superior influence for online learning course. 30% of the sample respondent did not agree that their supervisor will require online learning, 38% were neutral, and 32% believe that their supervisor might require it in future. The

Frequency

Frequency

  Std. Dev = . 

Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

positive response indicates that one third of the respondents think that in future they will have no choice since their supervisor will demand it, whereas a slight majority of the respondents provide neutral response indicating they are not sure whether their supervisor will need it in future. A third of the respondents do not believe their supervisor will require it in future. A survey of the owner operators may provide valuable insights and is suggested for future research. Figure 7 illustrates the histogram for the peer supervisor influence construct.

Efficacy 74% of the sample respondent agreed that they would feel comfortable using online learning,



Distance Learning in Business Aviation Industry

Figure 9. Histograms for the technology facilitating conditions construct Q

Q0

0

   0

0



Std. Dev = . Mean = . N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q

14% were neutral and only 12% assumed that they would not feel comfortable using online learning. The respondent average age was 43. We expected this generation of pilots to have much less confidence in using online learning. Since the great majority of the respondents assume that they would feel comfortable using online learning, we believe the niche market of business aviation pilots is more receptive for this product then previously believed. 60% felt confident that they could connect to online learning. 10% were neutral and 30% assumed that they would not feel confident. The confidence at being able to connect online indicates a high confidence and/or expectation of being able to use online learning anytime, anywhere. Figure 8 illustrates the histograms for the efficacy construct.

Facilitating conditions resources 42% of the sample respondents did not agree that there would be computer resource constraints for online learning, whereas 38% were neutral and 20% believe there will be resource constraints. The negative response clearly indicates that majority of the respondent do not think that they will face any resource constraint in using online courses. This may be interpreted as the pilots’ belief and expectation that their existing personal and work computers are adequate for online learning.



Frequency

Frequency

  Std. Dev = . 

Mean = .0 N = 0.00

0 .0

.0

.0

.0

.0

.0

.0

Q0

66% of the sample respondents did not agree that computer resources will be unavailable when they need to use them to take online learning, whereas 22% were neutral about it and only 12% believe there will have enough computer resources to get online when they need. The response to this variable indicates a confidence and /or expectation for anytime, anywhere online learning. Figure 9 illustrates the histograms for the technology facilitating conditions construct.

conclUsion This chapter explored the emerging opportunities and changing expectations contained in the vision of online learning for the business aviation pilot. It introduced and compared the concepts, perceptions and attitudes of online training for pilots and its relevance. The constructs efficacy (EFF), compatibility (COMP) and perceived usefulness (PU) are clearly seen as the top three determinants respectively of the pilots’ perceptions of what quality online training will provide. The responses to facilitating conditions resources (FCR) and ease of use (EU) constructs revealed a high degree of confidence in available resources and personal ability to perform online learning. The number of new pilot jobs created in the next decade is conservatively expected to grow at

Distance Learning in Business Aviation Industry

over 40%. In addition to planned growth, there is the need to replace aging business aviation pilots as they retire will also be significant. Training needs will therefore grow exponentially, as will the expectation by regulatory agencies and customers that future aviation training adopts the best of emerging technologies and techniques. The attitude of today’s business aviation pilot is changing rapidly. Two important factors that are pushing this change are first, the different training needs required by modern glass cockpit aircraft and second, the quickly changing awareness by the average pilot of new technologies, especially the Internet. Owners and operators of the business aviation industry will experience increasing financial pressure as globalization and substitute products provide cheaper aircraft charters for the customer and cheaper training for pilots. To remain competitive and maintain a leader in business aviation flight training, SimuFlite has an opportune position to combine their well-established training knowledge and techniques with the cutting edge technology of CAE to provide interactive distance learning. Pilots’ expectations of online learning are high but not overly demanding of current and emerging technology. More importantly perhaps are the pilots’ attitudes and beliefs brought out in the research that clearly shows trends in the acceptance of and confidence in online learning. It is very important that the online training experience be highly interactive with audio, visual, and tactile features. It must also be very flexible and easy to understand and use. Lessons should be modular and not only provide required training goals but also permit pilot(s) to explore deeper if desired. Real time and/or a virtual classroom environment were considered very important to business commercial aviation pilots. This not only permits them to feel that they are getting the most up to date and pertinent knowledge, but also provides an opportunity for networking of the pilots. The advantages of interactive online learning will be realized when broadband Internet, wire-

less protocol, and pilot familiarity converge and a critical mass of use and acceptance is achieved. For example, Khalifa and Liu (2008) discuss a conceptual facilitation form and knowledge acquisition using semantic network representation of computer-mediated discussions, and Hunter (2008) discusses a hybrid of service learning and Internet-based classes that may introduce a new motivation to e-learning while applying knowledge in a real world setting. Initially, to grow to the level of online training described above a business unit such as CAE SimuFlite can use current interactive technology such as Simfinity to offer lesson modules that also combine some audio, video, and textual training simultaneously. Before implementation however pilot reactions must be tested. Pilot interviews show that pilots are highly receptive to online training but have a “show me” attitude. To introduce too limited a product/service will likely turn pilots away and make it more difficult to attract them in the future. The business aviation industry is poised to undergo a fundamental change in the delivery of flight training to its pilots in order to accommodate cost and schedule pressures. The aviation training companies that intelligently embrace integrated distance learning to complement flight training will be able to not only provide a more personalized and differentiated product but also survive the inevitable technology driven changes in the global digital economy.

rEFErEncEs Azjen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action control: From cognition to behavior (pp. 11-39). New York: Springer-Verlag. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179-211.



Distance Learning in Business Aviation Industry

Biggs, J.B., & Collis, K.F. (1982). Evaluating the quality of learning – the SOLO taxonomy. New York: Academic Press. Butters, G. (1999, December). Learning At lightspeed. Business 2.0. Retrieved on February 16, 2003, from http://www.business2.com/articles/ mag/0,1640,13310,FF.html CAE SimuFlite. (2002, September 10). CAE SimuFlite Premiers Boeing Business Jet Training Program with CAE Simfinity™ Technology (Press Release). Retrieved February 19, 2003, from http://www.simuflite.com/press/02_11pr.html CAE SimuFlite. (2003, January 22). CAE to develop seLearning™ courseware for ALSTOM power plants (Press Release). Retrieved February 18, 2003, from http://www.cae.com/en/newsroom/2003/shtml/power_01222003_ref066. shtml CAE SimuFlite. (n.d.). Power systems and simulation. Retrieved February 19, 2003, from http:// www.cae.com/en/power/index.shtml CAE SimuFlite. (n.d.). CAE company overview. Retrieved January 28, 2003, from http://cae.com/ en/about/index.shtml CAE SimuFlite. (n.d.) CAE SimuFlite company history. Retrieved February 19, 2003, from http:// www.caesimuflite.com/press/history.html Cyber terror the way of the future. (1999, July 13). AAP General News (Australia). Retrieved February 19, 2003, from http://www.elibrary.com/ Davis, F.D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. Delio, M. (2000, August). Report: Online training ‘boring.’ Wired. Retrieved on February 16, 2003, from http://www.wired.com/news/ print/0,1294,38504,00.html

0

Dillon, H.C. (2007). Military training for civilian careers (Or: How to gain practical experience while serving your country). Occupational Outlook Quarterly, 51(1), 7-17. Greenburg, P. (2001). CRM at the speed of light. McGraw Hill. Herold, D.M., Davis, W., Fedor, D.B., & Parsons, C.K. (2002). Dispositional influences on transfer of learning in multistate training programs. Personnel Psychology, 55(4), 851-869. Hsu, S., & White, J. (1998). Developing holistic technology-based training programs. Retrieved February 17, 2003, from www.viats.org/papers/ session5a2.htm, 1, 2 Hunter, D. (2008). The virtual student/client experience. Journal of American Academy of Business, 12(1), 88-92. Karp, M.R., McCurry, W.K., Turney, M.A., & Harms, D. (2000). Aviation education for future airline pilots: An integrated model. Retrieved February 17, 2003, from www.viats.org/papers/ session5d4.htm, 1, 3 Karp, M., Condit, D., & Nullmeyer, R. (1999). F16 Cockpit/Crew Resource Management. NBAA Business Aviation Fact Book (2002) (AL/HR-TR1999-XXXX, in progress). Mesa, AZ: Air Force Research Laboratory.. Retrieved January 28, 2003, from www.nbaa.org/factbook/2002/section4.htm#02 Khalifa, M., & Liu, V. (2008). Semantic network representation of computer-mediated discussions: Conceptual facilitation form and knowledge acquisition. Omega, 36(2), 252. Mauro, R., & Barshi, I. (n.d.). Using an Internetbased decision research system in aviation training research. Retrieved February 17, 2003 from www.viats.org/papers/session5c2.htm Moore, G., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of

Distance Learning in Business Aviation Industry

adopting an information technology innovation. Information Systems Research, 2(3), 192-222.

Stokes, S. (2000, July 24). Learning space enlivens online training. Infoworld.

Leher, H.R., Moore, P.J., & Telfer, R.A. (1999). Approaches to learning: Is there a better way to prepare future pilots? Retrieved February 17, 2003, from www.viats.org/papers/session5d1. htm 1-2, 5

Taylor, S., & Todd, P. (1995). Understanding information technology usage: A test of competing model. Information Systems Research, 6(23), 144-176.

Rose, R.G. (2001, February 8). Practical use of pilot personality profile. Retrieved February 18, 2003, from http://www.avweb.com/news/ aeromed/181606-1.html Sand, K., & Schoenfelder, J. (1998). Simulation coupled with CBT creating a comprehensive training tool that increases transfer. Retrieved February 17, 2003, from www.viats.org/papers/ session5a4.htm, 1 Selvaratnam, S. (2002, November 11). Tool to create e-learning content. Computimes Malaysia. Shebilske, W.L., Jordan, J.A., Goettl, B.P., & Day, E.A. (1999). Cognitive and social influences in training teams for complex skills. Journal of Experimental Psychology: Applied, 5(3), 227-249. Small, R.L., Lakowske , S.D., Breese, J., & Callejo, G. (1999). A future direction in pilot training. Retrieved February 17, 2003, from www.viats. org/papers/session5d2.htm, 1-2 Sherry, L. (1996). Issues in distance learning. International Journal of Educational Telecommunications, 1(4), 337-365. Retrieved February 19, 2003, from http://carbon.cudenver.edu/~lsherry/ pubs/issues.html

Tomlinson, C.A. (1999, September). Mapping a route towards differentiated instruction. Educational Leadership. Trapp, R. (1997, September 3). Internet users ‘wide open” to fraud. Independent, p. C39. Retrieved February 19, 2003, from http://www. elibrary.com/. United States General Accounting Office. (1999). Aviation safety: Research supports limited use of personal computer aviation training devices for pilots (GAO Report No. RCED-99-143). Washington, DC: U.S. Government Printing Office. Retrieved on February 19, 2003, from http://www2. faa.gov/nsp/nsp/GAO_PCATD.txt FAA 1999. Wentling, T.L., Waight, C., Strazzo, D., File, J., La Fleur, J., & Kanfer, A. (2000, September). The future of e-learning: A corporate and an academic perspective. NCSA Knowledge and Learning Systems Group. Watson, J., Ahmed, P.K., & Hardaker, G. (2007). Creating domain independent adaptive e-learning systems using the sharable content object reference model, Campus - Wide Information Systems, 24(1), 45.



Distance Learning in Business Aviation Industry

aPPEnDiX a. sUrVEY QUEsTions Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

1. The online learning will be of no benefit to me. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

2. Using the online learning will improve my ability to get prepared for check ride. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

3. The advantages of the online learning will outweigh the disadvantages. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

4. Overall, using the online learning will be advantageous. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

5. Using the online learning will fit well with the way I take ground training. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

6. Using the online learning will fit into my work style. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

7. Instructions for online learning will be hard to follow. continued on following page



Distance Learning in Business Aviation Industry

aPPEnDiX a. conTinUED Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

8. Using online learning will require a lot of mental effort. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

9. Using online learning is often frustrating Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

10. I expect online learning would be clear and understandable Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

11. I expect online learning to be flexible to interact with. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

12. I expect it would be easy for me to become skilful at using online learning to get prepared for check ride. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

13. My friends would think that I should prepare for check ride through online learning.

continued on following page



Distance Learning in Business Aviation Industry

aPPEnDiX a. conTinUED Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

14. My fellow pilots would think that I should use online learning to get prepared for check ride. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

15. My supervisor would think that I should use the online learning to prepare for check ride. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

16. I will have to use online learning to conduct ground training because my supervisor may require it in the future. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

17. I would feel comfortable using the online learning on my own. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

18. If I wanted to, I could easily connect to get prepared for check ride through online learning on my own. Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

19. There will not be enough computer resources for everyone to use for online learning.

continued on following page



Distance Learning in Business Aviation Industry

aPPEnDiX a. conTinUED Strongly Disagree 1

Neither Agree Nor Disagree 2

3

4

Strongly Agree 5

6

7

20. I won’t be able to use computer resources for online learning when I need it.



Distance Learning in Business Aviation Industry

aPPEnDiX b. acronYms AERO AICC AT&T CBT CEO CRM DVD ERP FAA FMS FOQA HDTV HTBT IBM IDRS JAVA MGT NASA PRM SOLO U.S. WML XML



Aviation Education Reinforcement Option Aviation Industry Computer-based Training Committee Standards American Telegraph & Telephone Computer Based Training Chief Executive Office Customer Relationship Management Digital Video Disk Enterprise Resource Planning Federal Aviation Administration Flight Management System Flight Operations Quality Assurance High Definition Television Holistic Technology Based Training International Business Machines Internet-based Decision Research System Programming Language Developed for the Web Management National Aeronautics and Space Administration Partnership Relationship Management Structure of the Learning Outcome United States Wireless Markup Language Extensible Markup Language



Chapter X

SEAMAN:

A Visual Language-Based Tool for E-Learning Processes Gennaro Costagliola University of Salerno, Italy Filomena Ferrucci University of Salerno, Italy Giuseppe Polese University of Salerno, Italy Giuseppe Scanniello University of Basilicata, Italy

absTracT One of the crucial activities in the development of e-learning courses concerns the design phase. In this phase, instructional designers define the e-learning processes by specifying the activities students should carry out (knowledge objects, assessment, practice, etc.) and their temporal sequence. This phase is usually performed using an iterative process, with step-by-step refinements. Thus, it can greatly benefit of the availability of tools that assist instruction designers to carry out their work. In particular, a rapid prototyping approach could be effectively supported if the tool is also able to automatically generate the courses starting from the supplied specification. Moreover, such a tool should also provide support for reuse. To fulfil these requirements, in this chapter we present a tool based on a suite of visual languages, which has been specifically conceived to support instructional designers in the definition and creation of learning processes. The use of visual languages is motivated by the success they have achieved in other contexts (e.g., software engineering) for the construction of suitable models that allows to focus only on the features of interest and to provide more effective descriptions and reasoning. The proposed

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

SEAMAN

suite of visual languages includes the learning activity diagram, which extends UML activity diagrams to make them suitable for modelling e-learning processes, the Self-Consistent Learning Object language used to define knowledge contents, and the Test Maker Language for specifying assessment and self-assessment tests. The visual languages have been then implemented in SEAMAN (System for E-Learning Activity MANagement), a system prototype conceived to support instructional designers in the design, the generation, and the deployment of e-learning processes.

inTroDUcTion E-learning or electronic learning is a general term used to refer to computer-enhanced learning. The most notable advantages of e-learning are flexibility, convenience, and the ability to work at your own pace. In particular, groups of students participate and complete coursework enjoying e-learning activities in accordance with their daily commitments, thus making e-learning a viable alternative for learners with disabilities or those that have other commitments such as family or work. One of the crucial activities in the development of e-learning courses concerns the design phase. In this phase, instructional designers define the e-learning processes by specifying the activities students should carry out (knowledge objects, assessment, practice, etc.) and their temporal sequence. The e-learning evolution proposes a good number of approaches and tools aimed at assisting instructional designers during the analysis, design, and delivery of instruction (Bruce & Sleeman, 2000; Campbell & Mahling, 1999; Designer’s Edge, 2003; Goodyear, 1997; Schar & Kriger, 2000; Vrasidas, 2002). Many instruction design approaches proposed in the literature are based on traditional pedagogical learning approaches, or on the object-oriented paradigm. Indeed, existing models of instruction design have been influenced by linear or object-oriented software development processes. Nowadays, the new trend consists of exploiting ideas and benefits of component-based approaches for implementing and delivering learning environments. In particular, the idea is



to compose an e-learning process reusing learning components or activities, at different granularity levels (Rosenberg, 2001). In this chapter we describe a visual languagebased approach aimed at supporting the definition of e-learning processes assembling predefined didactic contents. The learning contents can be broken down and structured into a hierarchy from smaller, lower order blocks of material to higher, more complicated levels of learning. In particular, we have identified three different granularity levels referring to the size of knowledge contents. The use and assembling of these knowledge components provides the instruction designer with a modular paradigm to create distance courses, which resembles software development processes based on visual languages (Ferrucci et al., 2002; OMG Group, 1993). Hence, it has been defined a hierarchy of three visual languages to be employed during the different phases of the distance courses design process. Based on these languages we have constructed the System for E-Learning Activity MANagement (SEAMAN) to provide automated design support. The system and the underlying approach are particularly suitable for learning methodologies centred on didactic materials and assessment rules. The first visual language we propose extends the activity diagrams of UML (unified modelling language) (OMG Group, 1993) to enable the specification of didactic contents, assessment activities, and their relationships. For that reason such diagrams are named learning activity diagrams (LAD). They provide an explicit way to represent complex relationships between structural and

SEAMAN

behavioural e-learning activities. Any activity specified in a LAD sentence can be further refined by reusing previously defined e-learning activities or using a visual sentence belonging to either the self-consistent learning object (SCLO) language or the test maker language (TML). SCLO is a special case of state transition diagrams, and enables the instruction designer to define learning content objects. Instead, TML extends state diagrams to enable the design of assessment and self-assessment tests. It lets us describe tests that adapt themselves to student’s answers. It is worth noting that tests play an important role in our approach, allowing us to define learning processes adapting themselves to student performance. The chapter is organized as follows. The next section provides an overview of related work. Then the proposed visual languages are presented. The description of the SEAMAN system architecture, its facilities, and a sample application follows. Finally, a discussion on the achieved results and future work concludes the chapter.

rElaTED WorK Many activities regarding the e-learning process development are today accomplished by software tools that support instructional designers in their job (Bruce & Sleeman, 2000; Campbell & Mahling, 1999; Designer’s Edge, 2003; Goodyear, 1997; Vrasidas, 2002). In particular, the e-learning evolution proposes a good number of tools assisting instructional designers during the analysis, design, implementation, and delivery of instruction via the Web (Bruce & Sleeman, 2000). If, on one side, an automated support should be provided by authoring tools (Campbell & Mahling, 1998; Chang et al., 1996; Kasowitz, 1997; Thomson & Cooke, 2000), on the other side these tools should implement suitable e-learning process design methodologies (Douglas, 2001; Goodyear, 1997; Muraida & Spector, 1997; Vrasidas, 2002).

Muraida and Spector (1997), and Kasowitz (1997) review much of the work done in automated instruction design support tools. In particular, Muraida and Spector (1997) assert that there is “a lack of instructional designer expertise, pressure for increased productivity of designers, and the need to standardize products and ensure the effectiveness of product.” Thus, tools supporting instruction design during all the phases of the learning process definition are desirable. Goodyear views the instruction design as falling within four main approaches (Goodyear, 1997). These approaches allow the instruction designer to generate e-learning activities from given specifications by means of tools supporting the design of course structure, the selection of presentation templates, the reuse of design elements, and the coordination of activities accomplished by a design team. Moreover, Goodyear also proposes an approach for analyzing and designing distance courses that is divided into neat parts (Goodyear, 1999). The first part of Goodyear’s approach resembles the work of other people (outside education) who are interested in the design of technology supporting the work of information systems designers, requirements engineers, human factors specialists, and so on. The second part is instead focused on the design of good learning tasks exploiting traditional analysis and design processes. Often, these tools are not able to compensate the lack of expertise of instruction designers. Jones et al. (2003) have presented an information systems design theory for the design of information systems to be used in Web-based education. Vrasidas (2002) presents and discusses a system to develop hypermedia approaches as part of courses and learning environments delivered on the World Wide Web. It details the structuring of information, branching and interactivity, user interface, and navigation through Web-based distance courses. Opposed to these approaches, which are based on traditional models of instruction design, there are approaches and tools relying on object-oriented models. Douglas (2001) proposes an instruction



SEAMAN

design methodology based on the object-oriented paradigm. Designer’s Edge (2003) provides another interesting approach and tool for instruction design based on the object-oriented paradigm. Differently from the approaches we discussed above, AIMS (2004) Project describes a theoretical framework in which the knowledge domain editing and the course editing are distinguished. First the instructional designer constructs the domain model in terms of concepts and links. Finally, he/she defines the course structure starting from that description. Similarly, Thomson and Cooke (2000) propose the APHID method to support designers during the course creation by using instructional patterns. They also provide patterns describing teaching strategies known to be successful in particular situation. These patterns are used to design hypermedia applications. New trends seek means to exploit ideas and benefits of component-based approaches for implementing and delivering learning environments. In particular, the idea is to reuse learning components, at different granularity levels. At the topmost level there are existing self-consistent learning contents that may be composed of learning objects, which in turn may be composed of raw contents (Rosenberg, 2001). It is worth noting that the self-consistent learning materials can be seen as a framework in which instruction designers insert learning contents and raw contents specifying their interrelations and dependencies. Therefore, it could be interesting to reuse a selfcontained learning content and possibly also the associated learning objects. Lin et al. (2002) suggest the use of workflow technology to define and manage the coordination of e-learning activities. In particular, the authors introduce an e-learning environment, called FlexeL, which has been built upon workflow technology. The workflow functionality of Flex-eL manages the coordination of learning and assessment activities of the course process between students and teaching staff. In particular, this environment provides a unique environment for teachers to

0

design and develop process-centric courses and to monitor student progress. A process-modelling tool called FlowMake is also proposed in order to define e-learning processes. A process model is defined as workflow graph containing tasks and workflow modelling structures. Tasks are associated with roles and applications. The course activities and associated roles are identified and modelled using FlowMake. Personalization of the learning processes according to learners’ diversities is not provided. Differently, Carchiolo et al. (2002) present a prototype of a Web-based e-learning environment through which students can follow dynamically adapted learning process. In particular, the environment provides students with all formative paths moving from an initial to a desired knowledge, and where paths are adapted according to the student needs and capabilities, and dynamically modified according to the learners’ and teachers’ feedbacks.

VisUal langUagEs Visual and diagrammatic representations play a central role in several application domains, since they provide important tools for describing and reasoning. As visual languages have been applied to new application domains, such as spatial databases, education, software engineering, and so on, many different types of visual notations have been devised. In particular, in the software engineering domain they are widely employed for supporting the phases of the development process, such as requirements specification, analysis, and design. The numerous analogies between the software development and the instruction design processes suggested us to exploit visual languages to support several tasks of the instruction design process. Thus, the three visual languages we propose extend or resemble languages that have been successfully and largely used in the software engineering field to design software systems.

SEAMAN

The main language is the learning activity diagrams (LAD), which extends UML activity diagrams with means to model the workflows of learning processes (OMG Group, 1993). LAD can be used to model e-learning activities composed of distance modules, assessment, and self-assessment tests. States in LAD are activities, and most of the transitions are implicitly triggered upon the completion of the actions associated to the e-learning activities. The second visual language we have defined is used to refine the learning content objects, whereas the third one is employed to design the assessment and self-assessment tests. These visual languages have been introduced to refine the basic objects used within learning activity diagrams. They are named selfconsistent learning content objects (SCLO) and test maker language (TML). We describe them in the following subsections.

learning activity Diagrams The purpose of the LAD language is to model workflows associated to distance educational processes. As consequence, it allows instruction designers to describe educational materials, dependences, and assessment rules. Material dependences allow the author to vary the degree

of control over the order in which the students must explore the materials spread in SCLO objects. Moreover, using the results of assessment or self-assessment tests, the flow of the learning process is adapted to the learner performance. For example, before taking up a course the instruction designer can define a student test whose result may be used for assessing the knowledge, and to properly adapt the student learning process. The visual language symbols are shown in Figure 1. The first symbol (Figure 1A) represents a SCLO object, or content object for short. The name of the object can be placed in the symbol. This symbol represents a state of a learning process that is left when the associated learning object is completely executed. Every learning object can in turn be separately analysed and refined by using another visual language (Figure 1B). When this happens the icon shows a nested structure. The arrow (Figure 1C) represents the transition symbol, and it can contain a label. When the transition is not labelled, the only result of interest is content object completion. In those cases where it is important to know which content object has been completed, we associate different coloured coins to content objects execution. The synchronization symbol (Figure 1D) is a thick horizontal bar, and is used to coordinate content objects. The actions

Figure 1. LAD icons: A) SCLO object; B) refined self-consistent learning content object; C) transition element; D) synchronization bar element; E) self-assessment element; F) assessment element; G) merge element; H) start and stop marker



SEAMAN

underlying content objects may be concurrently executed more than once. The number of concurrent invocations is determined at runtime by a concurrency expression. The synchronization bar provides a simple way to express concepts like waiting for concurrent content objects to finish before proceeding forward along the learning process (join), and the starting of several content objects in parallel (fork). It is worth noting that by removing the synchronization bar elements from the LAD language we derive a special case of flow diagram, but with a considerably reduced language power. In fact, we cannot describe activities without dependences; hence all the activities have to be consumed in a sequential fashion. In the definition of processes focused on flows, and driven by internal processing, like for example industrial, didactic and software processes, this type of behaviour is vital. Figure 2. A LAD sentence



Let us now introduce two symbols describing assessment and self-assessment activities. The assessment test symbol (Figure 1F) is used to represent an activity aiming to evaluate the learner knowledge. The self-assessment activity (Figure 1E) is meant to be accomplished by the student to assess his/her knowledge. As a consequence, we differentiated the notations for these two symbols. Both symbols are used to represent decisions. As an alternative to guards on separate transitions leaving the same state, the aim of these objects is also to synchronize the incoming activities. These have one or more incoming arrows, and one or more outgoing arrows. The guard conditions are used to indicate different possible transitions that depend on test results. A decision may be shown by labelling multiple output transitions of an action with different guard conditions. These guard conditions may depend on self-consistent

SEAMAN

learning content objects that the instructor has to assess. We have used a merge symbol (Figure 1G) to merge back decision branches. A merge has two or more incoming arrows and one outgoing arrow. As opposed to the synchronization bar, the incoming transitions are not synchronized. Content experts, instruction designers, instruction technologists, media developers, and evaluation specialists are all professional figures that could be involved in the distance learning development process. For this reason, learning objects may be organized into swim lanes; the lanes can be used to organize learning objects with respect to these professional figures. The last two symbols (Figure 1H) are the start and the stop markers. They are used to indicate the initial and final states of a diagram. Figure 2 shows an example of LAD visual sentence, which represents a set of e-learning activities that have been structured into 4 swimlanes based on the learner knowledge. The student knowledge is assessed using the first test, so if the student has a score less than 60% then his/her knowledge is considered elementary. Thus, the learner has to study in depth the content presented in Material_1, and Material 2 before going on. Vice versa, when the test result is between 60% and 80% the contents presented to the learner will be those of Material_4. Finally, Material_3 is presented to the learner with advanced knowledge. After that, there are two parallel e-learning activities, which do not depend on the student knowledge. Following Material_7 there is an assessment test, so if one student has a learning deficiency in that self-consistent e-learning activity then he/she must

revise it and repeat the associated test. This means that the self-assessment element needs memory to remember the test result.

self-consistent learning object language The SCLO language is a kind of state transition language. Before defining it, we have investigated several languages used in multimedia software engineering (MSE). However, most of MSE languages have turned out to be complicated, requiring high expertise to be used. However, our aim was to formalize a visual language to be easily used by the target user within the visual environment implementing it. Thus, we have first defined a graph language to describe multimedia contents of learning processes. Four icons have been used to define this language. These are the multimedia node, multimedia link, start and stop marker node. Multimedia nodes (Figure 3A) represent the educational content that will be presented to the student. Typically they are composed by one or more multimedia raw contents. Thus, this node can be an atomic element or it may be composed aggregating atomic multimedia contents. The atomic content elements can be single Web page or simple multimedia objects, for example, short movies, songs, jokes, images, simulations, and so on. The multimedia object can have one or more incoming arrows, and one or more outgoing arrows (Figure 3B). Two multimedia nodes can be joined through a multimedia link. This represents the fact that students can browse multimedia contents by crossing links connecting objects. The last two symbols are the

Figure 3. The visual language elements: A) multimedia node or knowledge fragment; B) multimedia link; C) start marker node; D) stop marker node

A)

Multimedia Object

B)

C)

D)



SEAMAN

Figure 4. A visual sentence representing a learning content object

Figure 5. TML elements: A) question node; B) start and stop marker C) multimedia symbol; D-E-F) joint lines

Figure 6. A visual sentence representing a student assessment process



SEAMAN

start and the stop markers (Figure 3C and 3D). These are used to indicate the initial and final state. The system prototype we present in the chapter implements this visual environment by presenting a predefined page layout to the lecturer that he/she can use for managing knowledge content objects. Figure 4 shows a visual sentence example of the SCLO visual language.

the student test structure. The last symbol (Figure 5C) is recommended to motivate answers. An example of visual sentence from the TML is depicted in Figure 6. It is worth noting that swimlane tags can be used to declare the score that the learner must achieve in order to pass the tests.

Test maker language

sYsTEm ProToTYPE

Student assessment and self-assessment is a critical task in the knowledge process (Cynthia et al., 2000; Safoutin et al., 2000). The literature proposes a wide range of authoring tools to construct tests. Often, these tools do not have an associated visual environment to describe the test structure and its contents. In this chapter we introduce the test maker language (TML), a visual language supporting teachers during the design and implementation of tests. The language provides means to describe tests that adapt their contents to student answers. TML has five different symbols and three link types. The symbols are: question, aggregation symbol, multimedia object, and start and stop markers. The three link types define transitions between language symbols. Each of them has a different colour. The instructor can use the question symbol to represent one question and its associated answers, or he/she may refine it by using symbol annotations. Annotation is performed by using visual sentences from the same language. Question symbols (Figure 5A) are grouped with respect to knowledge contents and swim lanes. These are also used to give an execution order to regrouped answers associated to knowledge contents; the swim lanes order is from left to right. Links are used to model answers to questions. Thus, when a language sentence has an any link (Figure 5D) it means that the following question does not depend on the answer. Conversely, the false and true links (Figure 5E and 5F) modify

In this section we present SEAMAN (system for e-learning activity management), a prototype based on the described visual language hierarchy. SEAMAN assists instructors during the specification and implementation of e-learning processes and their associated e-learning activities, supporting the delivery of instruction via the Web. The system integrates modules for several authoring activities, such as knowledge contents, assessment, and self-assessment tests. The system can be configured as a centralized application so that the instruction designers can share and reuse content objects at different granularity levels. To better understand the functionality of SEAMAN, let us consider the use case diagram in Figure 7 showing the relationships among use cases, instruction designer, and the framework used to deliver learning processes. In particular, two actors were identified, namely the instructional designer and the e-learning framework. The former actor can define the learning process flow of e-learning courses, the knowledge content objects, and the tests using SEAMAN tool. Once the definition of learning process flow and its elearning activities is completed, the course can be generated. The generation process releases instruction contents to be deployed via Web, and will be available using an e-learning framework, as for example E-World (Casella et al., 2007). Figure 8 shows the layered architecture of SEAMAN. It includes three visual editors, one for each presented visual language, an application logic layer, corresponding to the generation engine



SEAMAN

Figure 7. A SEAMAN use case diagram D esign C ourse and A ctivities

D eploy C ourse and A ctivities Instructional D esigner

E - Learning F ram ew ork

G enerate C ourse and A ctivitie s

Figure 8. The SEAMAN architecture

laD Editor

sclo Editor

Tml Editor

generation Engine

repository

module, and a repository. The LAD editor allows the instructional designer to describe and model the learning process flow based on knowledge content objects, assessment, and self-assessment tests, and their dependences. The knowledge contents and the tests are defined by using the SCLO editor, and the TML editor, respectively. The generation engine module generates learning processes, knowledge contents, and the tests using



generation rules

the predefined generation rules stored within the repository of the system prototype. Furthermore, the generation engine module produces a zip file that contains didactic resources and configuration files. The didactic resources are composed of HTML files and SVG (scalable vector graphics, 2007) files, which are described by metadata files. Some configuration files to allow the e-learning framework to manage the learning process are also

SEAMAN

Figure 9. SEAMAN package diagram

graphic symbols

graphic Panel

input & output

graphical User interface (gUi)

Properties

Figure 10. SEAMAN high level class diagram overview



SEAMAN

produced. The generated e-learning course and its activities are stored in the repository to enable the instructional designer to eventually successively deploy them in an e-learning framework, for example, E-World (Casella et al., 2007). The SEAMAN tool is further described by the UML package diagram of Figure 9 and the UML class diagrams of Figure 10. In particular, Figure 9 shows the semantic dependencies among the packages: graphical user interface (GUI), graphic panel, graphic symbols, input & output, and properties. The GUI package is aimed at managing the graphical components, allowing the interaction between SEAMAN tool and the instructional designer. To insert and delete objects in the visual programming environments implemented in SEAMAN the classes in the graphic panel package have been implemented, while the classes in the graphic symbol package are aimed at visualizing the language objects in SEAMAN visual environments. In particular, this package contains the interface that all the visual objects have to implement since they can be managed by SEAMAN visual environments. The classes implementing the proposed visual objects are in this package too. The input and output of the visual sentences are managed by the classes of input & output package. Finally, the property package is used to manage the users’ preferences, and the multilingual menu of SEAMAN. The class diagram in Figure 10 shows the relations between the classes implementing the three visual editors and the generation engine. This class diagram highlights the extensibility and the flexibility of the proposed system prototype. It is worth noting that extensions or customizations of the proposed visual language hierarchy can be implemented by developers in SEAMAN with little effort. In such a way new types of e-learning activities can also be introduced to complete the formative offer. As output of SEAMAN is an e-learning environment that is delivered on the Web, features such as usability, colours, and graphical layout



become crucial for student welfare and e-learning course success. For this reason we defined several predefined graphical layouts that instruction designer chooses for the Web pages implementing the e-learning activities of a given process. Moreover, the prototype allows us to define new layouts or to customize existing ones. The e-learning activities generated by the prototype are iteratively navigable through a Web browser. Although the e-learning activities generated using the proposed approach are HTML pages, we need sophisticated technologies that only some browsers support. Thus, SEAMAN works for recent versions of Netscape and Internet Explorer. Although the language aims to support the design and development of learning processes, it has also turned out to be a powerful tool for presenting e-learning activities, and for monitoring student progress. Thus, an animation of diagrams allows students to monitor their progresses. Moreover, thanks to SVG storage format of LAD visual sentences, the student learning process can be visualized by using a Web browser with a suitable plug-in. The animation is executed only when the activities of a given distance course are deployed in an e-learning platform, implementing the learning traceability through a special conceived software component, such as the SCORM Run Time Environment (ADL, 2003). Moreover, SEAMAN generates a server module interacting with the LMS. The server module is integrated in the platform when the course is delivered. Colouring the activities green when the student finishes them, we provide a high level representation of the learning traceability.

a samPlE aPPlicaTion Several lecturers at University of Salerno have used SEAMAN to create e-learning courses. In particular, in this section we show its use for the design of the Programming Language

SEAMAN

Figure 11A. LAD sentence describing PLT learning process in SEAMAN

Figure 11B. LAD sentence describing PLT learning process displayed using Internet Explorer

Technologies course (PLT for short), belonging to the bachelor’s degree in computer science at University of Salerno. The aim of the PLT lecturer was to design the e-learning process to provide students with knowledge and expertise in the design and implementation of compilers. After attending the course generated starting from the defined e-learning process, the student should be able to design and implement a language compiler. In particular, as a course project the student is required to implement a subset of functionality of the Java language compiler back-end. The main steps to create a course with SEAMAN are:

1.

2.

3.

Use the LAD environment to define the course structure in terms of contents and tests; Use the SCLO environment to insert content for each SCLO object mentioned in the LAD sentence; Use the TML environment to define assessment and self-assessment tests in the LAD sentence.

Figure 11A depicts the visual sentence describing the learning process of the PLT course within the SEAMAN visual environment. On the other hand, Figure 11B highlights how the same sentence is presented to the student by using a Web browser with SVG plug-in. The sentence does not provide



SEAMAN

Figure 12. TML sentence describing the test on the runtime environment and compiler

Figure 13. A visual sentence representing the Compiler/Interpreter learning content object

0

SEAMAN

traceability animation because no e-learning activity has been completed. The two figures show that the course is divided in two parts. The former deepens topics about compilers preliminary notions, and basic tools for developing them (such as Lex and Yacc). The second part is conceived for designing and implementing compilers. In particular, the main characteristics of the JAVA, C#, and C++ compilers are shown. There are no dependencies among these topics, so their content can be also completed simultaneously. Starting from the objectives and the topics of the course the lecturer has introduced three tests, two of which are self-assessment and one is assessment. The first self-assessment test, depicted in Figures 11A and 11B, regards LEX and YACC tools. Since these tools are considered vital for the realization of the final course project, the lecturer has introduced a test and a feedback on them. Instead, the runtime environment and the compiler of the Java language are evaluated by using the second self-assessment test depicted in Figures 11A and 11B. The visual sentence describing this test is shown in Figure 12. In this picture we can also see an example of question with the associated answers. Finally, the lecturer has created a test for assessing the learners’ knowledge on the whole course contents. The visual sentence in Figure 13 shows the definition of the learning content object “Compilers/Interpreters” depicted in Figures 11A and 11B. The main fragment describes the compilation and interpretation processes, whereas subsequent fragments show the detailed activities.

conclUsion Academic and commercial e-learning authoring tools (Apple et al., 2002; Campbell & Mahling, 1998; Douglas, 2001; Goodyear, 1997; Muraida & Spector, 1997) use the basic concepts of the multimedia software engineering (MSE) (Christoffersen & Christoffersen, 1995). This is a new

frontier for both software engineering (SE) and visual languages (VL) (Bell & Jackson, 1992; Campbell & Mahling, 1998). As it has happened for these fields, also the e-learning field can benefit from the use of visual languages to simplify the work of an instruction designer. To this sake, in this chapter we have presented three visual languages for defining several stages of the e-learning course design process. The languages have been implemented within SEAMAN, a prototype that has been conceived to allow instruction designers to generate friendly learning environments and enhance their welfare. Experimental results have shown that the use of visual languages can encourage the design of distance courses, as opposed to what happened before with tools using more rudimental interaction paradigms. We also carried out a usability study of SEMAN using five lecturers at University of Salerno as subjects. The subjects used the SEAMAN system to produce a complete electronic version of their course. In particular, they have redesigned the following five courses from bachelors’ degrees related to computer science: Programming Language Technologies, Databases, Discrete Mathematics, Fundamental of Physics, and Business Administration. All the lecturers underwent an introductory course of six hours on the SEAMAN system and its visual notations. After that they were asked to use the system on their course, having the possibility to invoke individual tutor support. Once they had completed the design of their course with SEAMAN they were asked to fill in a questionnaire to provide feedbacks on system usability issues. In particular, other than the course name and the lecturer familiarity with diagrammatic languages, the form contained the following three categories of questions: intuitiveness of language symbols and visual sentences, training and usage times, and comprehensive tool evaluation with respect to well-known authoring tools. In general, we noted that familiarity with diagrammatic notations seems to facilitate tool



SEAMAN

usage. We also observed that non-computer science professors had more difficulties on the LAD language than on the other two visual languages. Probably, this was mainly due to the fact that although workflows should be familiar in business and many other disciplines, mapping the concept of activity synchronization on realworld problems is not immediate. Moreover, the discrete mathematics professor also had some difficulties on the definition of assessment and self-assessment tests with the TML language. More specifically, she found some problems in understanding the meaning of the true, false, and any joint links, and how they affect the test behaviour at run-time. In conclusion, the SCLO language was the easier to understand and use, whereas the LAD required some effort for people non-familiar with activity diagram notations. The hierarchical visual language organization appeared to be grasped quite intuitively by all of the lecturers. Regarding the time necessary to enter a visual sentence in SEAMAN this mostly reflected the familiarity with the given notation, although most of the lecturers took reasonable times. Finally, looking at the feedbacks of questions in the third group, it seemed that the use of visual languages in SEAMAN makes e-learning process creation somewhat easier as opposed to traditional authoring tools, and that all of the lecturers involved in this experiment expressed the possibly of using the SEAMAN prototype in the future. Future work will be devoted to extend the visual languages proposed in this chapter. This has a two-fold goal. On one hand, we aim to introduce less technical and more metaphor oriented visual languages to provide further abstraction in the e-learning design process, and consequently increase system usage among non-technical teachers. On the other hand, we aim to extend our visual languages in order to enable the modelling of additional aspects of the e-learning process. Special interest deserves the specification and the



realization of collaborative activities, in order to encourage cooperative and problem-based learning (Laister & Koubek, 2001). In this way groups of students can work together to solve problems, while keeping their diversities. As a result, the learning environments will encourage individual accountability, prompt feedback, and high selfexpectations.

rEFErEncEs Advanced Distributed Learning (ADL). (2003). SCORM present. Retrieved from http://www. adlnet.org AIMS: Adaptive Information System for Management of Learning Content. (2004). Retrieved from http://wwwis.win.tue.nl:8080/AIMS Apple, D.K., Nygren K.P., Williams, M.W., & Litynski, D.M. (2002). Distinguishing and elevating levels of learning in engineering and technology instruction. In Proceedings of 32nd ASEE/IEEE Frontiers in Education Conference, (pp. 6-9). Boston. Bell, M.A., & Jackson, D. (1992). Visual author languages for computer-aided learning. In Proceedings of IEEE Workshop on Visual Languages, (pp. 258-260). Seattle, WA, USA. Bruce, L.R., & Sleeman, P.J. (2000). Instructional design: A primer. Greenwich, CT: Information Age Publishing. Campbell, J.D., & Mahling, D.E. (1998). A visual language system for developing and presenting Internet-based education. In Proceedings of IEEE Symposium on Visual Languages, (pp. 66-67). Nova Scotia, Canada. Carchiolo, V., Longheu, A., & Malgeri, M. (2002). Adaptive formative paths in a Web-based learning environment. Educational Technology & Society, 5(4). From http://ifets.ieee.org/periodical/vol_4_2002/carchiolo.html

SEAMAN

Casella, G., Costagliola, G., Ferrucci, F., Polese, G., & Scanniello, G. (2007). A SCORM thin client architecture for e-learning systems based on Web services. International Journal of Distance Education Technologies, 5(1), 13-30. Chang, C.K., Chen, G.D., Liu, B.J., & Ou, K.L. (1996). A language for developing collaborative learning activities on World Wide Web. In Proceedings of 20th International Conference on Computer Software and Applications Conference, Seoul, South Korea (pp. 548-552). Christoffersen, R.D., & Christoffersen, A.H. (1995). Instructional system design: Its role in Coast Guard training and how it can influence the development of training materials. In Proceedings of the IEEE Professional Communication Conference: Smooth Sailing to the Future, Savannah, GA, USA (pp. 39-43). Cynthia, J., Finelli, M., & Wicks A. (2000, May). An instrument for assessing the effectiveness of the circuits curriculum in an electrical engineering program. IEEE Transactions on Education, 43(2), 137-142. Designer’s Edge. (2003). The industry standard instructional design tool. Retrieved from http:// www.allencomm.com/products/authoring_design/designer/ Douglas, I. (2001). Instructional design based on reusable learning object: Applying lessons of object-oriented software engineering to learning system design. In Proceedings of the ASEE/IEEE Frontiers in Education Conference, Reno, NV, USA (Vol. 3, pp. F4E-1-5). Ferrucci, F., Tortora, G., & Vitiello, G. (2002). Visual programming. In Encyclopaedia of Software Engineering. New York: John Wiley & Sons. Goodyear, P. (1997). Instructional design environments: Methods and tools for the design of complex instructional systems. In S. Dijkstra, N. Seel, F.

Schott & R. Tennyson (Eds.), Instructional design: International perspectives (pp. 83-111). Mahwah NJ: Lawrence Erlbaum Associates. Goodyear, P. (1999). Seeing learning as work: Implications for understanding and improving analysis and design. Journal of Courseware Engineering, 2, 3-11. Jones, D., Shirley, G., & Lynch, T. (2003). An information systems design theory for Web-based education. In Proceedings of the IASTED International Symposium on Web-Based Education, Greece. Kasowitz, A. (1997). Tool for automating instructional design. ERIC Educational Reports. Retrieved from http://ericit.org/digests/EDO-IR1998-01.shtml Laister, J., & Koubek, A. (2001). 3rd generation learning platforms requirements and motivation for collaborative learning. European Journal of Open and Distance Learning. Retrieved from http://www.eurodl.org/materials/contrib/2001/ icl01/laister.htm Lin, J., Ho, C., Sadiq, W., & Orlowska, M.E. (2002). Using workflow technology to manage flexible e-learning services. In Educational Technology & Society, 5(4). Retrieved from http://ifets.ieee. org/periodical/vol_4_2002/lin.html Muraida, D.J., & Spector, J.M. (1997). Automatic design instruction. In S. Dijkstra, N. Seel, F. Schott, & D. Tennyson (Eds.), Instructional design: International perspectives (Vol. 2). Mahwah, NJ: Lawrence Erlbaum. OMG Group. (1993). OMG Unified Modeling Language Specification. Retrieved from http://www. rational.com/media/uml/post.pdf Rosenberg, M. (2001). Mixing apples and oranges: Quick tips for surviving the interoperability myth. E-Learning Magazine, 2(10), 30-31.



SEAMAN

Safoutin, M.J., Atman, C.J., Adams, R., Rutar, T., Kramlich, J.C., & Fridley, J.L. (2000). A design attribute framework for course planning and learning assessment. IEEE Transactions on Education, 43, 188-199.

Thomson, J.R., & Cooke J. (2000). Generating instructional hypermedia with APHID. In Proceedings of the Eleventh ACM Conference on Hypertext and Hypermedia, San Antonio, Texas, USA (pp. 248-249).

Schar, S.G., & Kruger, H. (2000). Using new learning technologies with multimedia. IEEE Multimedia, 7(3), 40-51.

Vrasidas, C. (2002). A systematic approach for designing hypermedia environments for teaching and learning. International Journal of Instructional Media, 29(1). Retrieved from http://www. cait.org/vrasidas/hypermedia.pdf

Scalable Vector Graphics (SVG). (2007). XML Graphics for the Web. Retrieved from http://www. w3.org/Graphics/SVG/





Chapter XI

An Architecture for Online Laboratory E-Learning System Bing Duan Nanyang Technological University, Singapore Habib Mir M. Hosseini Nanyang Technological University, Singapore Keck Voon Ling Nanyang Technological University, Singapore Robert Kheng Leng Gay Nanyang Technological University, Singapore

absTracT Internet-based learning systems, or e-learning, are widely available in institutes, universities, and industrial companies, hosting regular or continuous education programs. The dream of teaching and learning from anywhere and at anytime becomes a reality due to the construction of e-learning infrastructure. Traditional teaching materials and methods are shifting to the new paradigm. In higher education, laboratory work is playing an important role in the area of training students and helping students to absorb more knowledge. With the goal of bringing e-learning to the traditional laboratory experiment, in this chapter, we present an architecture for an online laboratory e-learning system to facilitate the design and deployment of lab-based courses for e-education. The chapter provides an overall view of the system design and implementation so the Internet-based laboratory can be easily integrated with the e-learning infrastructure.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

An Architecture for Online Laboratory E-Learning System

inTroDUcTion The Internet and Web-related technologies are affecting more and more person’s lives and work around the world in many positive ways. They are also bringing many changes to the education domain. A major change in this area is the way educational materials are designed, developed, and delivered to the student. Internet-based learning, so-called e-learning, is widely available in institutes, universities, and industrial companies as regular or continuous education program, such as Blackboard (Blackboard Inc., n.d.), Swift Author (Gemini Inc., n.d.), Macromedia Authorware (Macromedia Inc., n.d.), and TopClass Publisher (WBT Systems, n.d.). Most of e-learning systems provide the services of searching, downloading, and delivering learning content, which includes text, audio, animation, applets, flash, or video clips to their users in order to enhance learning experiences. Internet-based laboratory (i.e., online laboratory) is a rapidly growing research in universities. In some cases, it also is called virtual lab, remote lab, Internet lab, or Web lab. In existing e-learning systems, the hardware equipment is not supported by the e-learning infrastructure, although IMS Learning Design Best Practice and Implementation Guide (IMS, ver1.0, 2003) presented some use cases of a virtual laboratory. These use cases have been chosen to validate the conceptual model of a learning system. However, it does not provide the details of design and implementation of such a lab-based learning platform in a systematic way. In this chapter, we have presented the architecture and modules of an online laboratory system, or OnlineLab for short. The main objective is to combine OnlineLab researches and e-learning infrastructure in order to achieve the goal of increasing and enhancing learning opportunities and experience for students. We first briefly present an overview of the e-learning system followed by a research review on Internet-based laboratories. Next, we provide details of the proposed architec-



ture for OnlineLab, which is based on the current Sharable Content Object Reference Model (ADL, SCROM) learning model. Finally, we demonstrate sample implementations based on the proposed model followed by concluding remarks.

oVErViEW oF E-lEarning sYsTEms The following equation represents a typical elearning system (Maish Nichani, 2001): LCMS = LMS + CMS [RLOs]

In this equation:

•

•

•

A Learning Content Management System (LCMS) is a “multi-user environment where learning developers create, store, reuse, manage, personalize, and deliver digital learning content from a central object repository” (elearningpost, n.d.). The main purpose of a learning management system (LMS) is to manage students and learning events and to collate data on learner progress. The objective of a content management system (CMS) is to simplify the creation and administration of online content (articles, reports, pictures, etc.) used in publications.

In a CMS, complete learning courses are assembled from several self-contained chunks called content components. These content components, when used in the learning domain, are called learning objects (LOs). One important benefit of the LO approach is reusability. Learning objects could be combined to form a hierarchy of lesson, module, course, or curriculum in order to provide a rich learning environment and to reduce the time, instructor skill, or cost associated with development. In this case, it is reusable learning objects (RLOs). With the LCMS, learners not only receive

An Architecture for Online Laboratory E-Learning System

the instructions when they desire (just-in-time learning) but also receive only the portion of the instruction that they desire (granular learning, or just-enough learning). Many organizations are working in one or more phases of the process in order to develop industrywide standards that ensure interoperability of learning solutions. Some are as follows:

•

•

•

•

Aviation Industry CBT Committee (AICC) was formed out of a need for hardware standardization of CBT delivery platforms in 1988. It has published a variety of recommendations, including hardware and software configuration. Their computermanaged instruction (CMI) guidelines have had the greatest impact. Advanced Distributed Learning (ADL) is an initiative launched in 1997 by the U.S. Department of Defense and the White House Office of Science and Technology Policy. Their recently released Shareable Courseware Object Reference Model (SCORM) provides one of the best and most recent examples of the application and integration of these learning standards. IMS Global Learning Consortium (IMS). Headquartered in Burlington, Massachusetts, it focuses on the development of XML-based specifications. These specifications describe the key characteristics of courses, lessons, assessments, learners, and groups. IEEE Learning Technology Standards Committee (LTSC) formed in 1996, is developing and promoting instructional technology standards. The most widely acknowledged specification is the Learning Object Metadata (LOM) specification, which defines element groups and elements that describe learning resources. The IMS and ADL both use the LOM elements and structures in their specifications.

rEViEW oF rEsEarch on ThE onlinE laboraTorY Benetazzo (1999) described the specification and design of a geographically distributed system based on commercial standard components. González-Castaño et al. (2000) designed an Internet access laboratory that provides remote access to real equipment on SBC68K—a singleboard computer based on a Motorola MC68000 microprocessor used in a Computer Architecture laboratory. Their system is based on object distribution paradigm in CORBA. A low-cost Internet-based telerobotic system was applied in remote robotic education via China Internet by Song You et al. (2000). Some issues involving time delays associated with the Internet also have been addressed. Chi and Chen (2001) conducted an experiment on a frequency modulation for students taking a course on communication principles at the National University of Singapore (NUS). Similarly, the REAL (Remotely Accessible Laboratory), a virtual laboratory accessible through the Internet, was implemented as a telematic service to allow remote control on mobile robots (Guimaraes, 2003). Professor Lambertus Hesselink (2003) from Stanford introduced a similar design for a remote laboratory, the so-called CyberLab, on an optical processor. Almost all proposed systems for an online laboratory basically provide remote access to the experiment for researching or teaching purposes. From the e-learning point of view, these systems do not include Learning Management System and Content Management System support. In other words, no universal platform has been developed so far in a way in which to provide a rich learning environment nor has there been proposed any standard development for designing and deploying lab experiment. However, the research works discussed previously provide a good ground for the design and implementation of an online laboratory system



An Architecture for Online Laboratory E-Learning System

from many aspects, such as system architecture, real-time issue, and Internet time delay. They also have presented a great deal of example applications widely used by universities, which can be better utilized to provided a more attractive and useful e-learning platform if some form of integration can be done.

hoW a lab EXPErimEnT is conDUcTED In this section, procedures for conducting a laboratory experiment are introduced. Table 1 shows a typical procedure for conducting an experiment in a real laboratory. We use the same procedure when we design our online lab. All the steps listed in Table 1 can be done via the Internet, if a proper interface and structure are defined. In this chapter, by traditional e-learning, we actually refer to the lessons, courses, or curriculum presented in text, audio, animation, applets, flash, or video clip format. They can run independently or can be downloaded from the LMS to the client machine. The significant difference between the online laboratory and traditional e-learning is that the former also contains the real hardware equipment. The equipment is not independent in such case and cannot be operated without the LMS’s Run-Time Environment

support. For a deeper analysis, the use cases of lab-based courses are discussed.

Use cases analysis Developing a model for a system prior to its construction or renovation is as essential as having a blueprint for a large building. The use case, borrowed from unified modeling language (UML), is adopted in this chapter. Similar to the real laboratory, lab-based e-learning courses must provide the same services for the student. Based on the IMS’s Learning Design Best Practice and Implementation Guide for a virtual lab, the use cases of lab-based courses in the e-learning domain are presented in Table 2. As the lab-based courses involve use of real hardware equipments, it is possible that unpredictable results or exceptions occur during a lab session. Table 3 summarizes some of the exception use cases. The proposed design for lab-based e-learning system courses is based on the use cases in Tables 2 and 3.

Table 1. Procedure of conducting laboratory experiment Stage



Procedure

Before the Lab-session

1). The lab-technician prepares the experiment (timetable, student’s list, apparatus, etc); 2). Student locates the laboratory and finds the experiment;

During the Lab-session

3). Student start the experiment work; 4). Asks help from lab-technician or supervisor if any doubt or question;

After the Lab-session

5). Student completes the work and collect the experiment result; 6). Submits a report; 7). The lab-technician and supervisor mark student’s experiment work.

An Architecture for Online Laboratory E-Learning System

Table 2. Use cases in success scenarios Type Success Scenario

Description For sys-admin: a). Adding new experiment, b). Removing experiment, c). Failure detecting/recovering, d). System Logging. For lab-manager: a). Providing experiment courseware, b). Assisting student to conduct experiment, c). Maintaining the experiment, d). Developing new Lab experiments, e).Managing students account. For students: a). Selecting laboratory, b). System opens lab environment, c). New student joining in a lab session and observing the experiment phenomenon, d). Passing the control to the partner to conduct the experiment, e). Completing lab activities, f). Closing the lab session, g). Completing learning objective.

Table 3. Use cases in exception Type

Description

Exceptions

For sys-admin: • System does not have experiment resources available. In this case the system must: Provide an alternative route to another similar equipment or, Display some relevant document. For lab-manager: • The student meets problem and leaves lab session in incomplete state. The lab-manager must: Provide the synchronise assistance by instant chat message or online phone calling, Reply the student by asynchronies method, i.e., Email, BBS. For students: •Students lost the connection or the experiment equipment is in improper running status. Then: Student closes lab session, System preserves lab environment for the student to continue the experiment, Student continues with other activities within the learning path.



An Architecture for Online Laboratory E-Learning System

ProPosED archiTEcTUrE For onlinE laboraTorY E-lEarning sYsTEm

Figure 1. Architecture for online laboratory elearning system

In this section, we introduce the architecture of the proposed online laboratory. Figure 1 shows an overall system architecture. The architecture is based on the SCORM specification for an e-learning system with additional module (Apparatus-LMS), which deals with functionalities for hardware-based learning systems. The SCORM specification by ADL deals with the launching, communicating, and tracking of content between the learning resources and the learning management system. It provides means by which learning resources can be reusable and interoperable across multiple LMS/LCMS systems. It consists of the following three components:

•

•

•

Learning Resources. Represent Assets (Web page, JavaScript, XML document, Flash object, picture, etc.) and sharable content object (SCO) (a collection of one or more assets). LMSAPIs. The communication mechanism between LMS and SCO. These are used for collecting and logging of learning-related data. SCROM Run-Time Environment. A Learning Management System that manages students and learning events to collate data on learner progress.

apparatus lms modules To extend the LMS to the lab-based courses, the proposed Apparatus LMS adds the following modules:

•

Apparatus Virtual User Interface. AppVUI for short, the remote control panel for the student to control the real apparatus and

•

•

0

observe the experiment status during the lab session. Apparatus Run-Time Environment. App Run-Time for short, the standard and uniform apparatus’ LMS environment, which provides laboratory services for students. Apparatus APIs. AppAPI for short, the communication mechanism for App Run-

An Architecture for Online Laboratory E-Learning System

Figure 2. Disgram of SOAP message

•

Time Environment and apparatus to exchange data. They provide all the necessary functions for initiating a connection to the actual apparatus and for delivering data and control commands. Web-Enabled Apparatus. The real hardware equipment, which consists of both the hardware and software that controls the hardware.

The most important element of the architecture is the App Run-Time Environment. It provides many standard functions, including lab session management, load balancer, report generator, and multi-user collaboration subsystem. These functions are reusable and shared by all learning objects, which are the experiments in the lab-based e-learning scenario.

morE DETails oF ThE archiTEcTUrE In this section, we present the details of the architecture modules.

apparatus Virtual User interface (app-VUi) To carry out an experiment from the Internet, students need to work in an interactive environment.

The App-VUI provides multiple students with a real-time environment to observe and control the experiment. The App-VUI allows the student to issue control commands through the user interface. Meanwhile, the experiment feedbacks, such as the apparatus response and status, are all shown in the App-VUI display area. To start a lab session, the App-VUI sets up an HTTP connection with the App Run-Time Environment using SOAP (Simple Object Access Protocol) protocol (W3C’s SOAP, n.d.). Once the connection is authorized, the student starts to conduct the experiment. Communication Using SOAP. SOAP is a lightweight protocol for exchange of information in a decentralized environment. It defines a base communication protocol in order for clients to exchange XML messages with the server. Figure 2 shows the diagram of a SOAP messaging binded with HTTP. In the proposed system, after the connection is established, the student issues the control commands, such as changing the parameter, reading the apparatus status, and downloading the experiment data. In most cases, the interaction between the student and LMS is a two-phase (request-response) message exchange. The student issues a request message, and the App Run-Time Environment replies with a response message. The response message carries back the status of the initial request (success or failure).



An Architecture for Online Laboratory E-Learning System

Table 4. Status code and description Status Description Code Normal Condition 0 No Error, successful operation execution Communications failure 101 Connecting with LMS failed 102 Apparatus is offline 103 Disconnected with apparatus 104 D isconnected with LMS 105 Sending data time out 106 Receiving data time out … Can be extended Data Errors 201 Invalided data in some form of format 202 Invalided data in some form of value … Can be extended

The following example shows a SOAP Message of Query the Apparatus Running Status. In this case, a GetApparatusRunStatus SOAP request message is sent to the LMS. Example 1. SOAP Message to Query the Apparatus Running Status

isRunning

The following feedback is an example response SOAP message containing the apparatus running status flag. Example 2. SOAP Response Message with the Apparatus Running Status Flag





Running

The information exchange between the user and the App Run-Time environment is based on XML-Messaging. This will enhance the readability for both the human being and the machine. Meanwhile, it is fully extendable for different experiments in different LMSs. XML-Based Messaging. XML is a markup language for documents containing structured information. Using XML-based messaging, the App Run-Time environment will be interoperable with any XML-based lab equipment. Thus, the lab experiment can be integrated into any LMS in an easy plug-and-play fashion. To make sure that the LMS server has received the message, a confirmation will be sent back to the client. In such case, a series of confirmation status codes is proposed. Table 4 shows the status codes we used in our system.

An Architecture for Online Laboratory E-Learning System

app run-Time Environment As discussed in the previous section, the App Run-Time Environment is the key to achieve the e-learning goal in lab-based courses. Figure 3 depicts the proposed model. There are three types of interfaces for the App Run-Time Environment to communicate with other modules:

•

•

•

Message Processor. Responsible for interacting with the student. Its functions include receiving the control commands and giving feedback on the experiment data and status to the student. AppAPIs. Provides communication channel for App Run-Time Environment and Apparatus to exchange data. These APIs provide all the necessary functions for initiating a connection to the actual apparatus and delivering data and commands. LMS Interface. Used to exchange information about the student and learning progress.

The App Run-Time Environment can be considered as a container (or framework) for the laboratory learning objects (LLO). Every LLO

that is integrated into the environment can share the resources provided by the framework. Apparatus APIs. Defined in ADL SCORM Run-Time Enironment (Version 1.3, 2004), a common mechanism for learning resources to communicate with an LMS and a predefined language or vocabulary form the basis of the communication. From the specification, only one communication scenario was proposed, which is Learner ⇔ LMS. However, the lab-based courses have one more type of communication: LMS ⇔ Apparatus. It is a dual communication in such case. To design a reusable and interoperable labbased experiment across multiples LMSs, there must be a common way to start, stop, and control the apparatus. Table 5 defines the App Run-Time Environment’s communication APIs in detail. Normally, the LMS provides the Web interface in order for the student to select the online experiment. Then, the App Run-Time will connect to the apparatus that the student selected via AppAPIs. Once the App Run-Time is connected to the apparatus, then the student can issue the control commands and receive the experiment data via the established communication socket. These AppAPIs fulfill the requirements for reusability and interoperability. They provide a

Figure 3. Proposed app run-time environment



An Architecture for Online Laboratory E-Learning System

Table 5. Apparatus APIs definition API

Execution

LOInitComm

Description: LMS initialises the communication with the Apparatus. It must be executed before calling other APIs. Parameter: The Apparatus’s system ID that are used to establish the connection with the apparatus. The ID is unique in such case. Return Value: Long integer indicates the connection handler if success. Otherwise, return 0.

LOCloseComm

Description: LMS closes the connection handler. Parameter: The return value of LOInitComm. Return Value: Boolean. “True” indicts the close action was successful. “False” means some error occurred when close the connection.

LOSendCmd

Description: LMS sends the command to the apparatus. Parameter: The return value of LOInitComm, Command String (defined in next section) Return Value: Boolean. “True” indicts the send action was successful. “False” means some error occurred when sending the command.

LOGetStatus

Description: LMS gets the status of apparatus, i.e., running status, apparatus’s indicator status. Parameter: The return value of LOInitComm, Status Code (defined by the different apparatus) Return Value: String. The status of the apparatus.

LOLastError

Description: This function returns an error code resulting from the previous API call. The code can be retrieved many times and will be kept unchanged until the new API call is made. Parameter: The return value of LOInitComm. Return Value: A pre-defined integer number, Error description string. More information refers to the XML-based Message

standardized method for an LMS to communicate with real lab experiments. Apparatus Queue. Considering the management of a real laboratory, all students who enrolled in the course share the lab resources by a scheduled timetable. In this chapter, the following two basic and important principles are refined:

• •

For a real lab session, only one user is permitted to conduct a particular experiment at a time. Any user can conduct a particular experiment in a specified time period.

Based on these principles, we have proposed a first-come-first-served apparatus queuing system with a time-out feature to ensure it is working in the App Run-Time Environment. Any user who wants to conduct a particular experiment must join the apparatus queue. For system identification, a unique UID corresponding to the user will be generated. After the first user who is conducting



the experiment is timed-out or terminated, the next user in the queue can start the experiment. The design guaranteed that only one user has the chance to access the apparatus at a time. Meanwhile, an unauthorized user cannot access control of the apparatus due to absence of UID. Session Management. In most cases, the lab technician is responsible for guiding the student to conduct the experiment and to collect the experiment results. In the App Run-Time Environment, the session management is acting in the same role. Sometimes, it provides synchronous assistance (by instant chat message or online phone calling) for the student when the important experiment is in progress and the student has a problem. On the other hand, the asynchronous assistance (by e-mail or BBS) also is included in some unimportant situation. To ensure high availability of lab-based courses to as many students as possible, the session management module is responsible for monitoring the lab sessions in order to do the following:

An Architecture for Online Laboratory E-Learning System

• • •

Terminate the experiment once time-out occurred Reset the apparatus for a new user Inform the lab manager in case of any hardware failure

Load Balancing. In the real world, lab resources, such as lab schedule, opening hours, equipment, and lab assistants, are always limited. Although lab resources now can be accessed 24 hours a day, 7 days a week through e-learning infrastructure, lab equipment is still limited. Hence, some form of load balancing is necessary. The purpose of load balancing is to increase utilization and enhance apparatus availability. On the other hand, high availability can be defined as redundancy. In this case, a few lab apparatuses that have the same functions and I/O need to be developed and made available in the App RunTime Environment. So, when several requests come to a same experiment, the system can redirect the request to point to a different apparatus according a prescheduled rule. Also, the whole e-;earning system can offer different services for special users, such as researchers and LO testers. It may become a useful and special feature for some commercial e-learning provider. Report Generator. Once the student completes the experiment, he or she may need to download the experiment results and control command data for further analysis or report. In general, these data are all ASCII-based. To enhance the readability for human being and machine, the proposed report generator processes these data according to the different XML schema. Multi-User Collaboration. When conducting the online experiment, students may feel isolated from both the teacher and their classmates, especially when the students have some questions about the experiment or have problems continuing with the experiment. Hence, the App Run-Time Environment must have a mechanism for remote

collaboration in order to allow students to conduct the experiment collaboratively. The benefits for the students working together in such a remote environment are the development of teamwork and the enhancement of the learning experience. Collaborative environment is achieved by including conferencing facilities, forum, and chat services to the system. Once the instructor initiates a lab session, many students can join the session. They can monitor the lab experiment by receiving the live video streamed by a camera that is pointed to the actual apparatus. At the end of the experiment, all the students in the same lab session can download the experimental results. More Modules. Besides the previous considerations, more useful features can be developed to enhance proposed model usability. Live video, discussion forum, Q&A whiteboard, and e-notification (by e-mail) can be embedded into the App Run-Time Environment. For the convenience of users, a new feature—SMS (short message service) reminder—also is introduced. SMS reminder provides an easy way to remind the students who is in the queue and when the apparatus is idle. Additionally, the administrator can get information about the laboratory by SMS in the area without Internet access. Normally, two ways can be employed to achieve SMS function via either SMS gateway (Internet) or GSM (Global System for Mobile Communication) modem. The second way is adopted in our scheme for the future system upgrade and migration.

Web-Enabled apparatus Current approaches to design online laboratory courses mainly are divided into the following two groups (González-Castaño, 2001):

• •

Use of educational simulators, and Remote access to real laboratory equipment.



An Architecture for Online Laboratory E-Learning System

In the first case, software simulations are used to simulate the behavior of real hardware apparatus. In the second case, the real hardware apparatus is used to allow the student to interact with the equipment. The proposed e-learning environment can work with both groups. Basically, the lab experiments are treated as a black box in both cases. These black boxes just provide the interface to communicate with the e-learning system. All data (i.e., apparatus status, experiment feedback) are transferred via the interface. A Web-enabled apparatus is a real hardware with a controller (i.e., electrical device, medical equipment, optical experiment) that can be connected to the network via TCP/IP protocol. The controller can be any form of embedded or PC-based device with a communication module that enables the connection to the network. For the software part, it executes local control of the hardware and also collects the experiment’s data. As a whole unit, the controller must have the physical interface, such as DA/AD converter or data acquisition card, in order to link with the actual equipment. Pastor (2003) presented an XML-based framework, REmote Laboratory Extended (RELATED), for the development of Web-based laboratories. The idea is to define an abstract entity called RLAB system by an XML DTD so that lab experiments can be described by RLAB. Then, the RLAB schema is published on the server for general use. An interesting research (Nacimiento’s VIML [Nacimiento, n.d.]) proposed a Virtual Instrumentation Markup Language (VIML), which is used to describe location, protocol, and device information for a network of virtual instrumentation devices and/or systems.

sample application In the proposed system, XML is used to describe the structure and attributes of the Web-enabled apparatus. As shown in Figure 4, a coupled tank apparatus has been developed. It consists of two



Figure 4. Coupled tank

small perspex tower-type tanks mounted above a reservoir that functions as storage for water. Two independent pumps pump water into the top of each tank. The apparatus is designed for teaching elementary feedback control principles. A FieldPoint (FP2000) running LabVIEW (National Instruments Corporation, n.d.) functions as the controller for the coupled-tank apparatus. The objective of this experiment is to maintain water levels in the tanks at the specified heights. From the control point of view, it can be treated as a Multi-Input-Multi-Out (MIMO) system (in this case, it is two-inputs [pumps] and two outputs [sensors] plant). The definition of the coupled tank is described by XML shown as the following: Example 3. XML Definition of the Coupled Tank Apparatus

... Coupled Tank Socket 192.168.0.1 2020 app-cpt-01-00

An Architecture for Online Laboratory E-Learning System

Figure 5. Experiment on coupled tank via App-VUI in NTU’s onlinelab system

float 0 5 float 0 5 float 0 5 float 0 5 …

The XML definition file has the following four significant sections: section. Gives information of the hardware equipment, such as apparatus name, communication mechanism, and so forth. section. A unique number across all LMS systems (may not necessarily be unique across all LMSs, but at least in any one LMS system). In this case, it is dynamically generated by the LMS and will be used for identification, operation, and so forth. section. Defines the input variables to be used to accept the control commands. Each input can have an arbitrary list of associated properties, each with a value. section. Defines the output variables to be used to send the apparatus data. Each output can have an arbitrary list of associated properties, each with a value.

An XML parser module corresponding to a definition file was developed in App Run-Time environment. Hence, the laboratory experiment can be recognized easily and identified by the



An Architecture for Online Laboratory E-Learning System

e-learning system. The changing, modifying, or upgrading of the experiment can be done easily by editing the XML definition file. In a normal condition, the controller of the apparatus actually is listening to the request from a TCP port. Once the command is received, verification will be processed to make sure the format and value of the command are correct and legal. Otherwise, the command will be discarded. As we can see, it is easy to have personalized learning objects. They can provide a different learning content or sequence to the student according to his or her learning experience. This App-VUI allows the student to experiment with different controllers, such as Manual, ON/OFF, PID (proportion-integral-derivative) control, and to study the effect of adjusting the various parameter settings available in the controller. As this is a relatively slow process, the water levels and the control signals to the pump can be charted on the App-VUI as the experiment progresses. The student also can download all data collected by the (NTU’s OnlineLab) server for further analysis when the experiment has been completed.

main important feature of our system is that it easily can be interoperated with any e-learning system that follows the standards.

rEFErEncEs ADL. (2004, January). SCORM run-time environment (RTE), version 1.3. Advanced Distributed Learning. (ADL). (n.d.). Retrieved from http://www.adlnet.org Aviation Industry CBT Committee (AICC). (n.d.). Retrieved from http://www.aicc.org Benetazzo, L., Bertocco, M., Ferraris, F., Ferrero, A., Offelli, C., Parvis, M., et al. (2000). A Web-based distributed virtual educational laboratory. IEEE Transactions on Instrumentation and Measurement, 49(2), 349-356. Blackboard Inc. (n.d.). Retrieved from http://www. blackboard.com Chalmers University of Technology’s I-Lab. (n.d.). Retrieved from http://www.ic.chalmers.se/ilab/ elearningpost. (n.d.). Retrieved from http://www. elearningpost.com

conclUsion E-learning is having a significant and positive impact on education. In order to understand how theoretical knowledge can be applied to real-world problems, experimental practical exercises are essential. In this chapter, we have proposed the architecture of an online laboratory e-learning system. The purpose is to combine the researches in Internet-based laboratory and e-learning infrastructures. We have provided the details of the modules. The system is based on XML messaging, which provides extensibility and plug-and-play fashion. We also have provided one example of the implementation. We are now conducting further studies to improve the performance of the system in terms of delays, efficiency, and scalability. The



Gemini Learning Systems Inc. (n.d.). Retrieved from http://www.gemini.com/ González-Castaño, F. J., Anido-Rifón, L., ValesAlonso, J., Fernández-Iglesias, M. J., Llamas Nistal, M., Rodríguez-Hernández, P., & PousadaCarballo, J. M. (2001). Internet access to real equipment at computer architecture laboratories using the ava/CORBA paradigm. Computers & Education, 36(2), 151-170. Guimaraes, E., Maffeis, A., Pereira, J., Russo, B., Cardozo, E., Bergerman, M., et al. (2003). REAL: A virtual laboratory for mobile robot experiments. IEEE Transactions on Education, 46(1), 37-42.

An Architecture for Online Laboratory E-Learning System

Hesselink, L., Rizal, D., Bjornson, E., Paik, S., Batra, R., Catrysse, P., et al. (2003). Standford cyberLab: Internet assisted laboratories. Journal of Distance Education Technologies, 1(1), 22-39. IEEE Learning Technology Standards Committee (LTSC). (n.d.). Retrieved from http://ltsc. ieee.org IMG Global Learning Consortium (IMG). (n.d.). Retrieved from http://www.imsglobal.org IMS. (2003). IMS learning design best practice and implementation guide, version 1.0 final specification. Ko, C. C., Chen, B. M., Hu, S., Ramakrishnan, V., Cheng, C. D., Zhuang, Y., et al. (2001). A Webbased virtual laboratory on a frequency modulation experiment. IEEE Transactions on Systems, Man and Cybernetics, 31(3), 295-303. Macromedia Inc. (n.d.). Retrieved from http:// www.macromedia.com Massachusetts Institute of Technology’s iLab project. (n.d.). Retrieved from http://ilabserv.mit. edu/ilab/ Nacimiento Inc. (2001). Virtual instrumentation markup language. Retrieved from http://www. nacimiento.com/VIML

National Instruments Corporation. (n.d.). Retrieved from http://www.ni.com National University of Singapore’s Vlab. (n.d.). Retrieved from http://vlab.ee.nus.edu.sg Object Management Group’s Unified Modeling Language. (n.d.). Retrieved from http://www. omg.org/ Rensselaer Polytechnic Institute’s AIM-Lab. (n.d.). Retrieved from http://nina.ecse.rpi.edu/ shur/remote/ Senvid’s summit system: Remote laboratory. (n.d.). Retrieved from https://www.senvid.net/ University Graduate Center’s LAB-on-WEB. (n.d.). Retrieved from http://www.lab-on-web. com/ W3C’s SOAP. (n.d.). Retrieved from http://www. w3.org/TR/soap/ WBT Systems. (n.d.). Retrieved from http://www. wbtsystems.com You, S., Wang, T., Eagleson, R., Meng, C., & Zhang, Q. (2001). A low-cost Internet-based telerobotic system for access to remote laboratories. Artificial Intelligence in Engineering, 15(3), 265-279.

Nanyang Technological University’s OnlineLab. (n.d.). Retrieved from http://www.onlinelab.eee. ntu.edu.sg:8000/

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 2, edited by T. K. Shih, pp. 87-101, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).



0

Chapter XII

A Virtual Laboratory for Digital Signal Processing Chyi-Ren Dow Feng Chia University, Taiwan Yi-Hsung Li Feng Chia University, Taiwan Jin-Yu Bai Feng Chia University, Taiwan

absTracT This work designs and implements a virtual digital signal processing laboratory (VDSPL). VDSPL consists of four parts: mobile agent execution environments, mobile agents, DSP development software, and DSP experimental platforms. The network capability of VDSPL is created by using mobile agent and wrapper techniques without modifying the source code of the original programs. VDSPL provides human-human and human-computer interaction for students and teachers, and it also can lighten the teacher’s load, increase the learning result of students, and improve the usage of network bandwidth. A prototype of VDSPL has been implemented by using the IBM Aglet system and Java Native Interface for DSP experimental platforms. Also, experimental results demonstrate that our system has received many positive feedbacks from both students and teachers.

inTroDUcTion Digital signal processing (DSP) (Mousavinezhad & Abdel-Qader, 2001; Texas Instrument, n.d.) is

one of the most powerful technologies in the 21st century and is a growing subject area in Electrical, Computer Science, and other Engineering/Science disciplines. DSP is linked closely to our life and

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

A Virtual Laboratory for Digital Signal Processing

is widely applied in many fields such as telecommunications, robotics, consumer electronics, medicine, military, instrumentation, aerospace, and automobile. Each of these areas has developed a deep DSP technology with its own algorithms, mathematics, and specialized techniques. Although DSP is the trend of current technology development, the learning of DSP is not an easy task for novices. Not only the DSP hardware architecture but also the flexible and powerful instruction sets of DSP chips are difficult for students. Thus, fast and convenient CAI tools for DSP learning are necessary. However, most DSP learning tools are stand-alone. This kind of learning approach has only human-computer interaction and lacks human-human interaction (Dey, 2000; Dow, Lin, Shen, Lin, & Chen, 2002) such as teacher-to-student and student-to-student. In order to add human-human interactions, it is necessary to create network capability for DSP learning tools. A network-enabled DSP learning environment can support multiple users and allow them to interact with each other in order to increase their interests in learning DSP in any place and at any time via the Internet. In addition to the network capability, a DSP virtual laboratory should support the features of multimedia and multilevel usage. The multilevel usage means that the same learning materials can be organized in different ways to be used in a regular semester course, a short course, an introductory exposition, an advanced seminar, and so forth, and by people with different linguistic, cultural, and perceptual preferences (Arndt, Chang, Guercio, & Maresca, 2002). Through multimedia demonstrations, students easily can understand various DSP theories. We can use the multimedia technology to enhance an experimental environment for students. Furthermore, DSP course material should be organized in multiple levels so students can select DSP studying materials according to their abilities in order to reduce the frustrations when learning and to deepen their impressions about DSP.

This work designs, develops, and implements a Virtual DSP Laboratory—VDSPL— using mobile agent and wrapper techniques. The autonomous feature of mobile agents can be used in the virtual laboratory to substitute for a teacher’s behaviors and actions in a practical laboratory. Mobile agents could guide several groups of students in different places simultaneously. When a student needs to interact with the teacher, the virtual laboratory can dispatch a mobile agent to perform this function. For a student, the mobile agent can play a learning guide and arrange the learning activities in order to improve the learning efficiency in a virtual laboratory. The rest of this chapter is organized as follows. First, in the second section, we discuss the background materials and related work. The third section describes the system architecture of our work. The system implementation and prototype are presented in the fourth and fifth sections, respectively. The sixth section shows our experimental results. Conclusions are finally offered in the final section.

rElaTED WorK There are many research areas related to our work, including virtual laboratory, digital signal processing, mobile agent techniques, and wrapper concept. These topics are described in this section. Distance education can be done in a wide variety of styles via different learning models. The virtual laboratory is one of the important components for macro university architecture (Arndt, Chang, Guercio, & Maresca, 2002; Dow et al., 2002). Students are required to learn some courses through online experiments and simulations, and the virtual laboratory is provided in order for students to conduct course-related experiments and simulations via networks. Based on the equipment and user access in each experiment, laboratories can be classified into four types



A Virtual Laboratory for Digital Signal Processing

(Dow et al., 2002). The first type of laboratory is the practical lab, which is a traditional laboratory. The second type of laboratory is the remote lab, which uses physical experimental equipment and allows users to remotely access the equipment and instruments. The third type is the micro lab, which provides some virtual equipment and allows only local access. Traditional computer-assisted instruction (CAI) tools belong to this type. The fourth type is the macro lab, which consists of one or more micro labs and allows remote access through the Internet. Some Web-based learning environments (Chang, Wu, Chiu, & Yu, 2003) belong to this type. The virtual laboratory proposed in this work is a hybrid of the remote lab and the macro lab. The theorems of DSP use the mathematics and the algorithms to manipulate the signals (Gan, Chong, Gong, & Tan, 2000; Wu et al., 2001) after they have been converted into a digital form. Currently, there are some DSP electronics manufacturers (e.g., TI, Motorola, NEC, Analog Device) that develop their own series of DSP chips. For instance, TI developed a series of highperformance DSP chips called TMS320™DSPs. In the past few years, it has been a challenge for students to learn the DSP concepts, theorems, and algorithms without any auxiliary simulating/emulating tools in a lecture class. With the rapid technological changing, there are several powerful simulation software tools (e.g., MATALAB and MATHCAD) (Causen, Spanias, Xavier, & Tampi, 1998) in order for students to learn DSP. Through these tools, students cannot practice DSP experiments with DSP hardware; they only can simulate digital signal processing by the simulation tools. The mobile agent (Concepcion, Ruan, & Samson, 2002; Dorca, Lopes, & Fernandes, 2003; Dow, Lin, & Hsu, 2002; Lange & Oshima, 1998; Pham & Karmouch, 1998; Silva, Silva, & Delgado, 1998) is an emerging technology that can be applied in many fields, including electronic



commerce, personal assistance, secure brokering, distributed information retrieval, telecommunication networks services, workflow applications and groupware, monitoring and notification, information dissemination, parallel processing, and so forth. The use of mobile agents can bring several advantages (Lange & Oshima, 1998; Silva, Silva, & Delgado, 1998), including the reduction of the network traffic and latency in client/server network computing paradigm, protocol encapsulation, dynamic adaptation, heterogeneity, robust, and fault tolerance. In the past few years, there have been several contemporary mobile agent systems developed (Lange & Oshima, 1998; Silva, Simões, Soares, Martins, Batista, Renato, et al., 1999; IBM’s Java Aglet, n.d.), and two main categories of mobile agent systems can be identified: systems based on the Java language (e.g., Mole [Pham & Karmouch, 1998], Aglets [Lange & Oshima, 1998; IBM’s Java Aglet, n.d.], Odyssey, Concordia, and Voyager [Pham & Karmouch, 1998]) and systems based on scripting languages (e.g., Agent Tcl [Pham & Karmouch, 1998], Ara Tcl-based Ara, and TACOMA). One important problem we face when building a virtual laboratory is where to place an extra function for a stand-alone learning tool without knowing its source code. To be included in a virtual laboratory and used via networks, these stand-alone learning tools have to be modified. In our approach, we use the wrapper concept to implement our virtual laboratory. Wrapper (Dow et al., 2002; Sudmann & Johansen, 2001) is a technique that provides a convenient way to expand upon existing functions of an application program without modifying its source code. Wrappers intercept function calls, method invocations, and messages to the application software that they wrap, redirecting or doing pre- and/or postprocessing of input/output. Wrappers provide a way to compose applications from different parts. The fact that a mobile agent is wrapped should be transparent to other mobile agents in the system and, potentially, to the agent itself.

A Virtual Laboratory for Digital Signal Processing

sYsTEm archiTEcTUrE The system architecture of our virtual laboratory and the functions of each component of the system are described in this section.

system overview In our proposed framework, we use the mobile agent techniques to construct the VDSPL. There are four major components in our virtual laboratory: the mobile agent execution environments, mobile agents, DSP development software, and DSP experimental platforms, as shown in Figure 1. Figure 1 also shows the mobility of agents between teacher and student sides via agent execution environments. Various mobile agents are

designed to assist a teacher. The mobile agents on the teacher side can be dispatched to the student side to represent the actual teacher and to interact directly with the students. The development tools are stand-alone programs on the teacher and student sides. Furthermore, some middlewares are designed and wrapped into our agents in order to provide interactive functions and rules for mobile agents and the development tool. Figure 2 shows an overview of our virtual digital signal processing laboratory. There are students, teachers, and learning Web sites in the virtual laboratory. The teacher can dispatch teacher agents to the student sides, and teacher agents will assist students and collect the information about the students’ learning status. In our

Figure 1. System architecture Network Agents

Agents

Environment

Environment

Development Tool

Development Tool

Student side

Teacher side

Figure 2. Overview of the virtual digital signal processing laboratory



A Virtual Laboratory for Digital Signal Processing

virtual laboratory, we also have a DSP Web site that provides the various DSP learning materials for students.

system modules The system modules are shown in Figure 3, and the details are described in this section.

Mobile Agent Execution Environment A mobile agent requires an execution environment called the mobile agent execution environment (MAEE) (Lange & Oshima, 1998). This environment must be installed on the student and teacher sides in order to provide a necessary runtime environment for agents to execute. The environment’s basic facilities include mobility, communications, naming and location, and security. All mobile agents are received and executed in the environment, and we also regard it as an entry point or operating system for mobile

Figure 3. System modules



agents. Furthermore, four important roles exist in MAEE, including engine, resources, location, and principal. The engine serves as a workhorse or virtual machine for MAEE and mobile agents. The resources include networks, database, processors, memory and other hardware, and software services. The location typically can be written as an Internet protocol (IP) address and a port of the engine with an MAEE name attribute. Principals like agents that have the responsibility for the operation of MAEE. The MAEE is implemented by using the Java language. Therefore, MAEE is a Java application that runs on the Java virtual machine (JVM) and has the following good properties: platform independence, secure execution, dynamic class loading, multithread programming, object serialization, and reflection.

Mobile Agent The mobile agent is a principal role in the virtual laboratory. Different mobile agents, such as the guide agent, demo agent, learning agent, monitor

A Virtual Laboratory for Digital Signal Processing

agent, homework agent, and assessment agent, can be designed for our learning environment. A teacher can use various mobile agents to assist students to learn. A guide agent can be used to provide an interactive interface between the teacher and the student. On the teacher side, the guide agent provides various assisting functions for the teacher. On the student side, the learning agent has some predefined FAQ rules, and it will reply appropriate answers from a knowledge base when the students ask some common questions or when the user’s behavior matches certain rules. The monitoring agent could act as the teacher to monitor the student’s actions and learning status. The homework agent could act as the teacher to dispatch homework to the student and to record the student’s homework execution status. The assessment agent could give an assessment to check the student’s learning results and to provide different levels of assessment materials. The demo agent helps the teacher to demonstrate the steps of the

experiment and to allow the student to have an overview of the experiment.

Wrapper A wrapper is the key component that provides the communication function in our framework for the VDSPL. The wrapper provides system function calls and gathers learning platform actions and information. When the wrapper is running, the mobile agent can interact with the experiment platform via the interface provided by the wrapper agent. The wrapper agent includes a service library and can be regarded as a fixed agent. The union of a mobile agent and a wrapper looks just like a stationary agent. This union can be wrapped, creating an onion-like structure with a core agent in the center and one or more wrappers around it. From a user’s perspective, the wrappers are hidden. A wrapped DSP development tool looks like any other software application, as shown in Figure 4,

Figure 4. Wrapping concept from a user’s perspective

DSP Development Tools

ATP

Agent

Wrapper2 Wrapper1

Figure 5. Wrapping concept from a system’s perspective

DSP

I/O

Development Tools

Function Calls

Command

Result

Interception

API

Response

Software Agent

Wrapper Agent



A Virtual Laboratory for Digital Signal Processing

where ATP denotes Agent Transfer Protocol. However, from a system’s perspective, wrapper is the agent itself. As shown in Figure 5, a wrapper agent consists of two parts: I/O interception and Application Programming Interface (API). I/O interception is in charge of exposing the functions of a DSP development tool as a set of methods by intercepting its I/O and commands.

DSP Experimental Environment The DSP experimental environment contains two parts: hardware environment and software environment. In the DSP software environment part, the software is the DSP program development tool, which is an existing application software without network capability and provides a powerful integrated environment and several necessary analysis tools to develop DSP programs. The software makes it easier and faster to implement DSP programs using C as opposed to the assembly language. The software also includes the debugging and real-time analysis capabilities. Currently, there are many DSP software environments, such as Code Composer, Matlab, Altera, and so forth. In the hardware part, the DSP experiment platform adopts the digital signal processor from the DSP chip manufacturer. The hardware platform consists of a DSP emulator and debuggers, which can support the user in debugging the DSP program code through a standard parallel port or PCI slot. Through the integration of the software and hardware environments, we can develop, debug, modify, and execute our DSP programs.

imPlEmEnTaTion This section describes our system implementation. The mobile agent and learning platforms are presented first. Expanding the network capability for the virtual laboratory system is described



next. Then, agent models and online learning implementation are presented.

Platforms Our virtual laboratory system consists of two platforms: the mobile agent platform and the DSP experimental platform. These two platforms are installed on the teacher and student sides. The mobile agent platform is Aglets, which was developed by the IBM Research Laboratory in Japan. The Aglets Software Developer Kit (ASDK) requires the JDK 1.1 or higher to be installed and is the first Internet agent system based on Java classes. The ASDK provides a modular structure and an easy-to-use API for the programming of Aglets. The Aglets are Java objects and can travel from a host to another host via networks. The migration of Aglets is based on a proprietary agent transfer protocol. An Aglet that executes on a host suddenly can halt execution, be dispatched to a remote host, and resume execution. When the Aglet moves, it takes along its program code as well as the states of all the objects that it is carrying. The security mechanism of Java virtual machine and Aglet makes a host safe when receiving the Aglets data. The DSP experimental platform is composed of TI’s integrated development tools, CCStudio, Dmatek PRO-OPEN TMS320C542 DSP Controller, and PICE-DSP ICE 320C542 (DMATEK Co. Ltd., n.d.). CCStudio software is a fully integrated development environment and supports TI’s leading DSP platforms. It integrates all hosts and target tools in a unified environment, including TI’s DSP/BIOS™kernel, code-generation tools, debugger, and real-time data exchange (RTDX) technology to simplify DSP system configuration and application design. CCStudio also has an open architecture that allows TI and third parties to extend the IDEs functionality by seamlessly plugging-in additional specialized tools. Through the CCStudio, students can learn DSP from mul-

A Virtual Laboratory for Digital Signal Processing

timedia presentation of real-world signals and system theories. Dmatek DSP Controller is an experimental board based on TI’s TMS320C542 DSP chip and designed in order for users to realize the function of the DSP chip and its peripheral device. PICE-DSP ICE 320C542 is an in-circuit emulator for DSPs.

allow us to enforce encapsulation of mobility management that enhances reuse and simplifies aglet design. Furthermore, the traveling patterns include three traveling models, including Itinerary pattern, Forwarding pattern, and Ticket pattern. In our approach, we use the Itinerary pattern and Forwarding pattern.

network capability

agent models

The network-enabled VDSPL capability is implemented by using Aglet design patterns and the wrapper concept. Design patterns are reusable components that have been proven to be very useful in the object-oriented field in order to achieve good application designs. The wrapper concept is used to expand new capabilities for an existing tool without modifying the original source code. The implementation of wrapper concept uses Aglet design patterns and the Java native interface (JNI). The Aglets design patterns include traveling patterns, task patterns, and interaction patterns. We add the network capability for the virtual laboratory by inheriting the traveling patterns. These patterns can deal with various aspects of managing the movements of mobile agents, such as routing and quality of service, and they also

In order to remotely control VDSPL, we use the JNI to connect the Win32 API in the initializeInterface. The Java native interface and Visual C++ are used to bind the Win32 API such as the jni2c.dll dynamic link library (DLL). An interface is initialized between other mobile agents and VDSPL for the wrapper agent. Moreover, the wrapper agent can execute a doCommand function that can be called by other mobile agents to control and monitor VDSPL. The wrapper agent also can respond to the results based on a wrapper script. Figure 6 shows the trigger of Windows API using JNI. In our system, there are six mobile agents implemented, including guide agent, monitor agent, demo agent, assessment agent, homework agent, and learning agent. These agents are de-

Figure 6. Trigger of Windows API using JNI



A Virtual Laboratory for Digital Signal Processing

signed for the platform on the teacher side and the student side, and each mobile agent has a different capability. The guide agent, assessment agent, demo agent, and homework agent work in the foreground. Agents mentioned previously have user interface to allow the user to interact directly with the system. Other agents without the awareness of their existence by the user work in the background. There are three basic patterns for an agent: Aglet class object, wrapper class object, and guide class object. The Aglet class allows the mobile agent to execute in the Aglet agent execution environment. This object class provides VDSPL the network capability. The wrapper class object provides the mobile agent a way to interact with the wrapper agent. The guide class allows agents to communicate and to interact with the user. This class provides function calls for the wrapper script. Each type of mobile agent uses different teaching and learning knowledge-based rules. If the predicate of each rule is satisfied, the mobile agent will take predefined actions.

Figure 7. Guide agent



sYsTEm ProToTYPE A prototype of VDSPL is presented in this section. As shown in Figure 7, when the mobile agent platform starts running, it first will initiate an experimental platform and provide an agent list for the teacher. If the teacher needs an agent service or wants to communicate with students, the guide agent can be used to do so. Figure 8 shows a learning agent that supports different materials and topics. After the guide agent clones a learning agent for students, the learning agent will carry the learning materials that are determined by the teacher. When the learning agent starts, the students will receive a message informing them, and the learning program will start and then load the DSP learning materials. Sometimes, the teacher wants to provide a demonstration of the experiment for students in an experimental course. The learning agent can employ the guide agent to collect the student actions and experimental results and to dispatch a demo agent to students. After the guide agent clones a demo agent for students, the demo agent will carry a predefined script. If the students have problems, then the learning agent will notify the teacher’s guide agent. The teacher then will interact with the student through the guide agent.

A Virtual Laboratory for Digital Signal Processing

Figure 8. Learning agent

Figure 9. A snapshot of the system when the demo agent starts

Figure 9 is a snapshot of the prototype when the demo agent starts, and the demonstration example will be presented step by step according to the demo script. In VDSPL, we also have created a Web site that provides news, DSP introduction, DSP material zone, online learning, download, discussion board, and related links.

According to each student’s status, the teacher can assign an assessment to the students. The assessment agent supports different assessment levels and exercises, as shown in Figure 10. The students will be informed, the assessment program will start, and then an assessment is loaded. The agent will record automatically the scores or carry



A Virtual Laboratory for Digital Signal Processing

Figure 10. Assessment dispatching

the assessment results and send to the teacher’s guide agent when they finish their work.

EXPErimEnTal rEsUlTs Experiments were conducted and surveys were taken to evaluate the user satisfaction of our system. Graduate students and teachers in our department were recruited to conduct these experiments. A total of 15 graduate students and five teachers were inquired in our experiments. We investigated the user satisfaction of our system from the points of view of both students and teachers for the following system metrics, including demonstration, interaction, monitoring, assessment, and network capability. The demonstration function could demonstrate the steps of the experiment and let students have an overview of the experiment; this function is provided by the demo agent. The interaction function could provide an interface between a teacher and a student to interact with each other; this function is provided by the guide agent. The monitoring function could act as the teacher to monitor the actions and learning statuses

0

of students; this function is provided by the monitor agent. The assessment function could give an assessment to check a student’s learning results and to provide different levels of assessment materials; this function is provided by the assessment agent. The network capability is provided by using the wrapper and agent techniques to enable the network function of stand-alone DSP development tools. The feedbacks of students and teachers for the five system metrics are shown in Figures 11 and 12, respectively. We can observe that both students and teachers have positive feedback for our system, especially for the functions of demonstration and network capability. In addition to evaluating the user satisfaction of our system, experiments also were conducted to evaluate the importance of these five functions for a virtual laboratory of DSP experiments. From the students’ points of the view, the importance of these functions from high to low is demonstration, network capability, interaction, assessment, and monitoring. From the teachers’ points of the view, the importance of these functions from high to low is demonstration, monitoring, assessment,

A Virtual Laboratory for Digital Signal Processing

Figure 11. VDSP metrics from the students’ points of view Strongly Agree

Agree

5

4

3

Somewhat Agree

2 Oppose

1 Strongly Oppose

0

Demonstration

Interaction

Monitoring

Assessment

Network Capability

Figure 12. VDSP metrics from the teachers’ points of view 5 Strongly Agree

4 Agree

3 Somewhat Agree

2 Oppose

1

0

network capability, and interaction. We can observe that demonstration is the most importance function for a virtual laboratory from the points of view of both students and teachers. The reason is because demonstrating the steps of an experiment is very important for conducting a laboratory. From the teachers’ points of view, the function of monitoring is also very important. However, it is less important from the students’ points of view.

This is because a teacher may want to know the learning statuses and behaviors of their students. However, most of the students prefer a more carefree learning environment without the teacher to tie them down. As shown in Figures 11 and 12, it is also very interesting that the monitoring function provided by our system is enough from the students’ points of view, but it could be improved from the teachers’ points of view.



A Virtual Laboratory for Digital Signal Processing

conclUsion In this chapter, we present VDSPL, a mobile agent-based virtual digital signal processing laboratory. Our system incorporates agent techniques with DSP development tools to provide teachers and students with various instructions and interactions. The mobile agent and wrapper techniques are used to enable the network capability of stand-alone DSP development tools and to improve the teacher-to-student interaction for distance DSP learning. Furthermore, students can get guidance and learn in the personalized environment through mobile agents. In addition, the mobile agent and design patterns also are used to perform software re-engineering and to provide a virtual laboratory.

rEFErEncEs Arndt, T., Chang, S. K., Guercio, A., & Maresca, P. (2002). An XML-based approach to multimedia software engineering for distance learning. In Proceedings of the 14th International Conference on Software Engineering and Knowledge Engineering, (pp. 525-532). Ischia, Italy. Causen, A., Spanias, A., Xavier, A., & Tampi, M. (1998). A Java signal analysis tool for signal processing experiments. In Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, Vol. 3, (pp. 1849-1852). Seattle, Washington, USA. Chang, S. K., Arndt, T., Levialdi, S., Liu, A. C., Ma, J., Shih, T., et al. (2000). Macro university: A framework for a federation of virtual universities. International Journal of Computer Processing of Oriental Languages, 13(3), 205-221. Chang, W. F., Wu, Y. C., Chiu, C. W., & Yu, W. C. (2003). Design and implementation of a Webbased distance PLC laboratory. In Proceedings of the 35th Southeastern Symposium on System



Theory, (pp. 326-329). Morgantown, West Virginia, USA. Concepcion, A. I., Ruan, J., & Samson, R. R. (2002). SPIDER: A multi-agent architecture for Internet distributed computing system. In Proceedings of the ISCA 15th International Conference on Parallel and Distributed Computing Systems, (pp. 147-152). Louisville, Kentucky, USA. Dey, A. K. (2000). Enabling the use of context in interactive applications. In Proceedings of the 2000 Conference on Human Factors in Computing Systems, (pp. 79-80). Hague, The Netherlands. DMATEK Co. Ltd. (n.d.). Retrieved from http:// www.dmatek.com.tw Dorca, F. A., Lopes, C. R., & Fernandes, M. A. (2003). A multiagent architecture for distance education systems. In Proceedings of the 3rd IEEE International Conference on Advanced Learning Technologies (pp. 368-369), Athens, Greece. Dow, C. R, Lin, C. Y., & Hsu, F. W. (2002). A mobile agent-based virtual language learning laboratory. In Proceedings of the International Conference on Chinese Language Computing, (pp. 98-103). Taichung, Taiwan. Dow, C. R., Lin, C. Y., Shen, C. C., Lin, J. H., & Chen, S. C. (2002). A virtual laboratory for macro universities using mobile agent techniques. The International Journal of Computer Processing of Oriental Languages, 15(1), 1-18. Gan, W. S., Chong, Y. K., Gong, W., & Tan, W. T. (2000). Rapid prototyping system for teaching real-time digital signal processing. IEEE Transactions on Education, 43(1), 19-24. IBM’s Java Aglet. (n.d.). Retrieved from http:// www.trl.ibm.com/aglets Lange, D. B., & Oshima, M. (1998). Programming and deploying Java mobile agents with aglets. Boston: Addison Wesley.

A Virtual Laboratory for Digital Signal Processing

Mousavinezhad, S. H., & Abdel-Qader, I. M. (2001). Digital signal processing in theory and practice. In Proceedings of the 31st ASEE/IEEE Frontiers in Education Conference (pp. T2C-13T2C-16). Reno, Nevada, USA. Pham, V. A., & Karmouch, A. (1998). Mobile software agents: An overview. IEEE Communications Magazine, 36(7), 26-37. Silva, A., Silva, M. M., & Delgado, J. (1998). AgentSpace: A next-generation mobile agent system. Lecture Notes in Computer Science, 1477, 148-159. Silva, L. M., Simões, P., Soares, G., Martins, P., Batista, V., Renato, C., et al. (1999). JAMES: A platform of mobile agents for the management

of telecommunication networks. In Proceedings of the 3rd International Workshop on Intelligent Agents for Telecommunication Applications, Vol. 1669, (pp. 77-95). Stockholm, Sweden. Sudmann, N. P., & Johansen, D. (2001). Supporting mobile agent applications using wrappers. In Proceedings of the 12th International Workshop on Database and Expert Systems Applications (pp. 689-695). Munich, Germany. Texas Instrument. (n.d.). Retrieved from http:// www.ti.com.tw Wu, H. T., Hsiao, T. C., Chen, C. L., Su, C. M., Su, J. C., & Jiang, J. C. (2001). An integrated teaching and learning DSP lab. system. Journal of Science and Technology, 10(1), 29-36.

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 2, edited by T. K. Shih, pp. 31-43, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).





Chapter XIII

Information Retrieval in Virtual Universities Juha Puustjärvi Helsinki University of Technology, Finland Päivi Pöyry Helsinki University of Technology, Finland

absTracT Information retrieval in the context of virtual universities deals with the representation, organization, and access to learning objects. The representation and organization of learning objects should provide the learner with an easy access to the learning objects. In this chapter, we give an overview of the ONES system, and analyze the relevance of two information retrieval models for virtual universities. We argue that keywords based search (i.e., the Boolean model), though well suited for Web searches, is overly coarse for virtual universities. Instead, the vector model, on which our implemented search engine is also based on, seems to be more appropriate as it provides similarity measure (i.e., the learning object having the best match is presented first). We also compare the performance of four algorithms for computing the similarities (matching).

inTroDUcTion Today people in all professions are faced with increasing demands. Technology develops in an ever-increasing speed, and the roles of people in work, society, and industry are shifting constantly. Keeping up with the pace of change requires con-

tinuous education and learning. Traditional campus-universities are trying to answer to this need of lifelong learning by building virtual universities, whilst facing competition from the commercial continuing education providers in the form of e-learning.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Information Retrieval in Virtual Universities

E-learning can be defined as information technology enabled and supported form of distance learning, in which the traditional restrictions of classroom learning have disappeared. The main tool of e-learning is a personal computer, and the Internet serves as the principal communication and distribution channel. The learners can participate in online Web-based courses and interact with both the peers, instructors, and the learning materials. E-learning sets new requirements for universities: they have to build global learning infrastructures, course material has to be in digital form, course material has to be distributed, and learners must have access to various virtual universities. As single virtual universities are independently created, they may provide very heterogeneous functionalities and user interfaces. Ideally, the learner should be able to access all the virtual universities in a similar way (i.e., the heterogeneity of various virtual universities should not burden the learner). How this goal can be achieved is the main topic of the ONES-project. Consequently, the main functions of the ONES system are to hide the distribution of e-learning portals, and to hide the semantic heterogeneity (i.e., problems arising from using same words in different meaning and vice versa). In order to achieve these goals, the system will deploy many new technologies such as “onestop portals,” Web services, service oriented architecture, RDF-based annotation, ontology editors, and distance measures in searching learning objects. In this chapter, we will restrict ourselves on the role of searches in the ONES-system. In particular, we will analyze the applicability of different information retrieval technologies. Our main argument is that the technology based on the Boolean model (Yan & Garcia-Molina, 1994), though well suited for searches in the Web, is not suitable for the emerging virtual universities. Instead, for

virtual universities we have to develop methods, which allow learners to be more concerned with retrieving information about a subject than with retrieving data, which satisfy a given query. For example, a learner may be interested in courses dealing with object-oriented programming rather than in the courses where the term “java” or “C++” is stated. When searching for information about a subject (e.g., object oriented programming) the search engine must somehow interpret the metadata of the learning objects and rank them according to a degree of relevance to the learner’s query. The primary goal is to retrieve all the learning objects, which are relevant to a learner’s query while retrieving as few non-relevant objects as possible. Unfortunately, characterization of the learner’s information need is not a simple task. Furthermore, the difficulty is not only in expressing the information need but also in knowing how the learning objects should be characterized with the help of the metadata descriptions. The rest of this chapter is organized as follows. First, in the second section we give an overview of the architecture of the ONES-system. In the third section we characterize virtual universities. In particular, we will give an overview of the e-learning environment, and specify what the notion of resource-based learning incorporates. Then, in the fourth section, the role of metadata and ontologies in virtual universities is illustrated. In addition, the usability of the Boolean and the vector model in a virtual university is analyzed. Especially, two interpretations of a hierarchical ontology in the context of the vector model, called weighted leaves and multilevel weighting, are introduced. Then, in the fifth section, the performance of four matching algorithms based on weighted leaves and multilevel weighting principles is compared. Finally, the sixth section concludes the chapter by summarizing the feasibility of the proposed ideas.



Information Retrieval in Virtual Universities

ThE archiTEcTUrE oF ThE onEs sYsTEm The name ONES stands for one stop e-learning portal. As this name suggests, a salient feature of the system is the aggregation of distance learning information from different learning sources in one portal. The idea of the one-stop portals originated from one-stop shops, and later on it is also adopted in e-government applications. All one-stop applications have the same goal: hide the heterogeneity and distribution of local systems. So, from user’s point of view one-stop portal behaves like a centralized system. The four main components of the ONES-system are (see Figure 1): • • • •

Aggregation portal (mediator) Wrappers E-learning portals Course providers’ tools

The aggregation portal supports the learners in searching the courses that match to their specific needs. It differs from traditional database interfaces in a way that in addition to the traditional database queries it supports fuzzy queries. Fuzzy queries are similarity based, which means that if

the similarity between the courses’ profiles and the learner’s query exceeds a certain threshold, they are said to match. A problem is that the current database management systems do not support fuzzy queries and therefore the ONES-system has to support them. From technological point of view, the aggregation portal is a mediator (Garcia-Molina, Ullman, & Widom, 2000). It supports a virtual view that integrates several learning sources in much the same way as data warehouses do. However, since the mediator does not store any data, the mechanisms of mediators and warehouses are rather different. Since the mediator has no data of its own, it must get the relevant data from its sources and use that data to form the answer to the learner’s query. As the data sources (e-learning portals) are independently created it is obvious that they provide heterogeneous interfaces (e.g., they may provide different kind of functionalities or the same functionalities are provided by different operations). In order to hide this heterogeneity there is a wrapper (Garcia-Molina et al., 2000) between the mediator and each e-learning portal. So a wrapper is a software module that extracts data from local e-learning portals. This implies that the wrapper must be able to accept a variety of

Figure 1. ONES-architecture Lea rner

Learner Aggregation portal (a mediator)

Course provider



Wrapper

Wrapper

eLea rning portal

eLearning portal

Course provider’s tool

Course provider’s tool

Course provider

Information Retrieval in Virtual Universities

queries from the mediator and translate any of them to the terms of local eLearning portal. The wrapper must also communicate the result to the mediator. An important point is that each wrapper provides equal functionality for the mediator. Ideally, each wrapper provides an interface for requesting the metadata of learning objects (i.e., descriptive information of courses, course packages and programs offered by educational institutions, e.g., universities). From a technological point of view, each e-learning portal is a Web service (Vasudevan, 2001). Web services are self-describing modular applications that can be published, located, and invoked across the Web. Once a service is deployed, other applications (e.g., an aggregation portal) can invoke the deployed service. In general, a Web service can be anything from a simple request to complicated business process. A course provider can enter data about a course through the course provider’s tool. The main function of this tool is to provide an interface, which facilitates the creation of the metadata attached to learning objects. Basically, this tool is analogous to the tools that support the content providers of electronic newspapers (Yli-Koivisto & Puustjärvi, 2002) in creating metadata items to news articles. The tool may even generate suggestions of the suitable metadata items, after which the author can make the necessary modifications and enter this information to the system.

characTErisTics oF VirTUal UniVErsiTiEs E-learning Environment E-learning can be defined as information technology enabled and supported form of distance learning, in which the traditional restrictions of classroom learning have disappeared (Liu, Chan, Hung, & Lee, 2002). The main tool of e-learning is

a personal computer, and the Internet servers as the principal communication and distribution channel. The learners can participate in online Webbased courses and interact with both the peers and instructors and with the learning materials. The teacher-centeredness of traditional learning does not hold for e-learning, where the learning process has become more and more learner centered. The learning process and the resources may be customized according to the individual needs of the learner. At the same time, the role of the teacher becomes that of a facilitator or of a mentor guiding and supporting the individual process of learning (Liu et al., 2002). Typical e-learning environments, such as WebCT and Virtual-U, offer the basic elements for delivering e-learning courses: course content delivery tools, synchronous and asynchronous discussion forums and conferencing systems, possibilities for quizzes and polling, workspaces for sharing resources, white boards, possibilities for evaluation and grading, logbooks, possibilities for submitting assignments, and so forth (Liu et al., 2002).

studying in Virtual Universities In the recent years, the idea of a virtual university has been becoming more and more popular in many countries all over the world. The enormous development in the field of information and communication technologies has enabled the rise of e-learning and virtual learning environments. As a result, the traditional universities have faced a new challenge emerging from the commercial sector of education. There is a growing need for new kind of learning and teaching as the technology advances rapidly and the skills and competencies required in the working life become more demanding and increasingly dynamic. Virtual university has been defined as a space where the students are provided with higher education courses with the help of the newest



Information Retrieval in Virtual Universities

information and communication technology (Niemi, 2002). The degree of utilizing technology in organizing the studies may vary from pure technology-based studies to face-to-face or mixed studies that are supported by learning technologies. The main channel of communication and delivery of teaching is the Internet (Niemi, 2002; Ryan, Scott, Freeman, & Patel, 2000). Thus, a virtual university can be seen as closely related to e-learning that provides learning opportunities via the Internet. The difference between these two concepts is the level of studies offered; virtual university is aimed to offer higher education studies while e-learning can be used for all educational levels. A virtual university may be an institution that uses the information and communication technologies for its core activities such as providing learning opportunities, administration, materials development and distribution, delivering teaching and tuition, and providing counseling, advising and examinations. On the other hand, a virtual university may also be a virtual organization created through partnerships between traditional universities and other educational institutes. In addition, the traditional campus universities may be regarded as virtual universities if they offer learning opportunities via the Internet or combine traditional ways of learning with e-learning (Ryan et al., 2000). Virtual universities are expected to offer opportunities for life-long learning for audiences otherwise excluded from university studies. The emerging virtual university can be seen very beneficial especially for the industry, when technology-supported learning can be brought to the workplaces and integrated more closely to work. Moreover, virtual university can enhance organizational learning and bring competitive advantage by continuously developing the skills and knowledge of the employees (Teare, Davies, & Sandelands, 1999).



resource-based learning The Internet is able to store and transmit vast amounts of information in different forms and formats. Therefore the Internet is an ideal support for resource-based learning (RBL) that is one of the corner stones of learning and teaching in the virtual university. RBL has been defined a student-centered way of learning that exploits various specially designed learning materials, interactive media and technologies. RBL can be realized as self-study or as interactive group learning both in distance and in the face-to-face mode (Ryan et al., 2000). The Internet can be used to enable and support RBL in several ways (Ryan et al., 2000): • • • • • •

Courses can be delivered via the Internet. Resources can be identified and used. Internet serves as a communication and conferencing channel. Learning activities and assessment can be done in the Net. Collaborative work is enabled. Student management and support is enabled.

In the next section, we focus on the starting point of RBL, namely on searching learning resources.

inFormaTion rETriEVal moDEls Information retrieval in the context of virtual universities deals with the representation, organization, and access to learning objects. The representation and organization of learning objects should provide the learner with an easy access to the learning objects. The system retrieves all the learning objects, which are relevant to learner while retrieving as few non-relevant learning objects as possible

Information Retrieval in Virtual Universities

In this section, we will analyze the usefulness of different information retrieval models (Baeza-Yates & Ribeiro-Neto, 1999) for a virtual university. The used model determines the way the metadata of the learning objects are given as well as the way the learner’s queries (information needs) are presented. Before analyzing the information retrieval models we characterize the role of metadata and ontologies in virtual universities.

metadata and ontologies In order to transfer data seamlessly and efficiently in the virtual university, there has to be a standard way for both people and computers to communicate all necessary knowledge, with both people and computer systems (Stojanovic, Staab, & Studer, 2001). One possible solution is to use metadata and an ontology attached to it for describing the learning objects. The term metadata has variable interpretations depending upon the circumstances in which it is used. For example, in the context of documents the common forms of metadata include the author(s), the source of publication, the length of document, and so forth. This kind of metadata in commonly called descriptive metadata. For example, the metadata elements of the Dublin Core (Pöyry, Pelto-Aho, & Puustjärvi, 2002) represent descriptive metadata. Educational metadata is needed for improving the retrieval of learning objects, for supporting the management of collections of learning objects, and for supporting the decision process of the learners looking for educational resources. LOM seems to be the most powerful and most widely used metadata standard for educational information systems (Holzinger, Kleinberger, & Müller, 2001; Lamminaho, 2000). More generally, educational metadata can be used by educational institutes and professionals as well as by learners in order to describe (e.g., the content, structures, and relationships of the learning objects and to

search for educational objects) (Lamminaho, 2000; Stojanovic et al., 2001). Educational metadata may describe any class of educational objects, such as study courses. The pedagogical features of the course, the contents, special target groups, and the technical requirements of the study course can be described with the help of a metadata schema (Lamminaho, 2000). More generally, educational metadata can be used to describe, for example, the content, structures, and relationships of the learning objects (Stojanovic et al. 2001). Educational metadata can be utilized by educational and pedagogical professionals, by the institutions offering education, and by the students searching for education. Well-designed and sufficient metadata aid the decision making process of the students and help the educational institutions to provide suitable information about their educational supply (Lamminaho, 2000). Educational metadata is very much semantic metadata, but a thorough metadata schema must include also at least structural metadata in order to be able to describe the learning objects efficiently. The idea of using standardized metadata schemas is being able to develop universally applicable tools dealing with the metadata descriptions of the learning objects. In order to create metadata records containing the resource descriptions specific tools are needed for creating the metadata according to the standards (Kassanke, El-Saddik, & Steinacker, 2001). Metadata is also useful when guiding non-experienced users through a large collection of learning resources (Strijker, 2001). Moreover, metadata is seen as value-added information that is used to arrange, describe, track or otherwise enhance the access to the object content. At the moment metadata becoming increasingly important when digital government and e-commerce are emerging. Metadata enables increased accessibility, expanded use of objects, multi-versioning, and system improvement. The granularity of metadata, which refers to the level of details in the description, is an important ques-



Information Retrieval in Virtual Universities

tion when developing a metadata set (GillilandSwetland, 2000). A salient feature of descriptive metadata is that it is external to the meaning of the document, (i.e., it describes the creation of the document rather than the content of the document). The metadata describing the content of the document is commonly called semantic metadata. For example, the keywords attached to many scientific articles represent semantic metadata (Jokela, 2001). An ontology provides a general vocabulary of a certain domain (Fridman & McGuinness, 2001), and it can be defined as “an explicit specification of a conceptualisation” (Gruber, 1993). In essence, an ontology gives the semantics to the metadata. Ontologies are formal, explicit, and shared specifications of some conceptualizations. Formal means that the ontology should be machine readable, and explicit refers to having defined the types of concepts and the constraints on their use are explicitly defined. Shared refers to the fact that an ontology must reach a consensus (Fensel, 2001). Ontologies together with metadata enhance efficient access to information by offering possibilities to organize and categorize the content of the information system in question. In this context an ontology is defined as a means to formalize and to specify a common terminology for a defined area of interest (Turpeinen, 2000). In order to standardize semantic metadata specific ontologies are introduced in many disci-

plines. Typically, such ontologies are hierarchical taxonomies of terms describing certain topics. For example, the ACM Computing Classification System is a hierarchy (a tree) in which the nodes represent the classes of the taxonomy. In Figure 2, a subset of that hierarchy is represented.

The boolean model Applying the Boolean model in searches requires that each learning object is augmented by a set of metadata items such as keywords or classification identifiers (e.g., the searches in the CUBER system (Pöyry et al., 2002; Pöyry & Puustjärvi, 2003) are based on the Boolean model). A learner can then query learning objects by Boolean expressions comprising of operands and operations. The operands are the used keywords and the operators are typically “and,” “or,” and “not.” For example, by using ACM Computing Classification system (Figure 2) the keywords attached to a learning object might be D, H.1, and H.2.2 (corresponding the keywords Software, Models and Principles, and Physical Design). Now, if a learner presents the query “D and (B or H.1)” (i.e., learning objects having the keyword “Software” and at least one of the keywords “Hardware” and “Models and Principles”), then the previous learning object will match that query. The Boolean model is intuitive and clear. Moreover, it can be efficiently implemented even

Figure 2. A subset of the ACM Computing Classification System Subject

D. Software

H. Informat ion Systems

H.1. Models and Principles

H.2.1.Logical Design

00

H.2.2. Physical Design

B. Hardware

H.2. Database Management

H.2.3. Languages

H.2.4. Systems

Information Retrieval in Virtual Universities

in the case of huge amount of objects. For example, many Web search engines are based on this model. However, using that model in a virtual university gives rise to following drawbacks: •

•

•

First, the model is based on a binary decision criterion, meaning that each learning object is predicted to be relevant or non-relevant. In reality, it is obvious that the resulting learning objects fit more or less to the query (i.e., some kind of grading should be possible). Second, expressing the requirements of learning objects by a Boolean expression may be difficult. Third, a typical problem concerning search engines based on the Boolean model is that either the result of the query includes too many or too few learning objects.

In the next section, we consider a more advanced model, which avoids many of the drawbacks just described.

The Vector model The vector model differs from the Boolean model in that weights can be assigned to each metadata item of a document as well as to the keywords of the query. The idea behind this model is that we can more accurately specify the queries and

the contents of the documents (e.g., learning objects). Assuming that the standard metadata items (e.g., the classes in Figure 2) specify a vector space (i.e., each item (keyword) in the hierarchy represents a dimension in the vector space), we can represent each document and query as a vector in that vector space. Then we can process the query by computing the distance of the query vector and the document vectors. This kind of computing requires that the sum of the weights of each document and query equals to a predefined constant. For convenience, the used constant is usually one. As the result of the query the documents are sorted in the order determined by the similarity (i.e., the document having the best match with the query is presented first). The number of the documents in the result should be restricted by requiring a certain degree of similarity. Using the vector model in a virtual university requires that the course provider assign the metadata items and their weights into each learning object. The metadata items to be used are selected from the used domain ontology. Depending on the used course provider’s interface this can be done in various ways. For example, as in our prototype system, there may be an ontology structure on which the course provider inserts the weights. In Figure 3, the ontology structure of the Figure

Figure 3. A metadata specification of a learning object Subject

D. Software

H. Informat ion Systems

H.1. Models and Principles

H.2.1.Logical Design

H.2.2. Physical Design 0.1

B. Hardware 0.3

H.2. Database Management 0.6

H.2.3. Languages

H.2.4. Systems

0

Information Retrieval in Virtual Universities

2 is augmented by setting weights on the nodes “B.H.2,” and “H.2.2.” Note that the node having no weight means that its weight is actually zero. Hence, the profile of the learning object can be presented by a vector in 9-dimensional vector space as follows: [0 x D, 0 x H, 0.3 x B, 0 x H.1, 0.6 x H.2, 0 x H.2.1, 0.1 x H.2.2, 0 x H.2.3, 0 x H.2.4]. That is, the profile is a point in an orthogonal 9-dimensional vector space. The gain of attaching metadata description for learning objects is that we can use mathematical distance measures in computing learners’ queries. Further, computing the distance requires that the descriptions (vectors) be specified in an orthogonal vector space. In other words, the nodes in the hierarchy that are used in profile vectors must be independent. In practice this means that we have to follow one or the other of the following interpretations:

be an appropriate assignment. On the other hand, if the course is very specific then the weight of H.2 could be zero. If we follow the weighted leaves interpretation, then in determining the profile of a learning object weights are set only on the leave nodes of the hierarchy. Consequently, the profiles of the learning objects are specified by vectors in an orthogonal vector space, which is determined by the leave nodes of the hierarchy. To illustrate this approach let us consider the weighting of the course “Physical design in database management systems.” In this case, all the weights are given on the nodes H.2.2 (Physical design) and H.2.4 (Systems) independently of the level of the course.

•

The learner presents queries in the same way as the content provider determines the weights of the learning object; both these are presented by vectors. Hence the query presents an ideal profile of the learning objects that satisfy the learner’s requirements. For example, assuming that the multilevel weighting interpretation of the ontology is used, and a learner wants to find basic courses concerning database management. In this case the learner will set rather heavy weight on H.2 (database Management) and lighter weights on H.2.1 (Logical Design), H.2.2 (Physical Design) and H.2.3 (Languages). In contrast, if a student is looking more advanced courses on database management then the student will give a lighter weight on H.2 and heavier weights on its siblings. As the learners interact with the system by submitting queries it is reasonable to require that the response times should be only a few seconds. We investigated the effects of different matching algorithms and the amount of stored learning objects on response times. The test environment was equipped with Pentium II processor and 192 MB memory. The computers were running the Sun

•

Multilevel weighting interpretation: The leaves and the nodes of the ontology hierarchy represent independent concepts. Weighted leaves interpretation: The parent node represents the union of its siblings. In other words, each sibling represents a subset of its parent. Yet the siblings represent independent concepts.

The intuition behind multilevel weighting is that we can express the level of a leaning object (as well of a query) by altering the weights on a node and its siblings. To illustrate this let us consider the weighting of the course “Physical design in database management systems.” Now, it is obvious that the weights should be given on the node H.2 (Database management) and its siblings H.2.2 (Physical design) and H.2.4 (Systems). Assuming that approximately half of the course deals with databases in general and the other part deals with physical design and database management systems, then giving weight 0.4 to H.2 (Database management), 0.3 to H.2.2 (Physical design) and weight 0.3 to H.2.4 (Systems) could

0

ProcEssing lEarnEr’s QUEriEs

Information Retrieval in Virtual Universities

Solaris 5.8 operating system. We implemented and tested four matching algorithms (i.e., algorithms) which compute the distance measures of learning objects and learners’ queries. We next give a short description of the algorithms. The Cosine matching algorithm (Baeza-Yates et al., 1999) calculates the cosine measure between the query (a vector) and the documents profiles. As a matter of fact the algorithm does not compute distance measures but rather approximates distance measures by computing the angles of the query vector and the vectors representing documents, such as the learning objects. The Euclidean matching algorithm (Friedman, Bentley, & Finkel, 1977) calculates the Euclidean distance from the query profile to all learning objects’ profiles. The Manhattan distance algorithm (Bentley, Weide, & Yao, 1980) calculates a so called “city block-distance.” The name comes from the fact that this measure in two dimensions tells how many blocks in a city one would have to walk between two points. Our developed Fuzzy matching algorithm attempts to achieve more efficient matching procedure than the “exact” matching algorithms. The improved efficiency is achieved by performing the actual matching on a pre-selected subset of all learning objects. The predefined subset of the documents’ profiles is determined by choosing the three biggest weights from the query and then computing the subset based on these weights. Then only the profiles, the weights of which are within a specified tolerance interval are selected for the final query processing. Therefore the result set

is not guaranteed to contain all the profiles that are closest to the matching profile. However, the closeness values of the profiles in the actual result set are exact, since they are calculated using the Euclidean measure. The computing time for matching of each algorithm is presented in Table 1. The test was performed for different amount (1000, 5000 and 10 000) of learning objects. Basically, the differences of Euclidean, Cosine and Manhattan algorithms were rather small (less than 10%). Fuzzy matching algorithm required least computing time (about 20% less than others). However, the test proves that all the algorithms are quick enough in the test environment as the response times are less than 1.2 seconds. If the number of the learning objects or the dimensions of the vector space (i.e., the used attributes in the profile) increases, then it obvious that the Fuzzy Matching algorithm will be more superior to the other algorithms. In our test environment the vector space comprised of 15 dimensions (i.e., each profile could have at most 15 attributes). In practice, the number of attributes cannot increase significantly as otherwise the determining the weights for learning objects would overly burden the coarse creators. In addition, as the system is developed for universities it is not obvious that number of learning objects can be very huge (e.g., over 10,000).

Table 1. Matching times for the algorithms 1 000

5 000

10 000

Manhattan

0.80 0.83 0.83

0.93 0.96 0.98

1.07 1.16 1.23

Fuzzy

0.61

0.73

0.89

Cosine Euclidean

0

Information Retrieval in Virtual Universities

conclUsion Virtual university has been defined as a space where the students are provided with higher education courses with the help of the newest information and communication technology (Niemi, 2002). The degree of utilizing technology in organizing the studies may vary from pure technology-based studies to face-to-face or mixed studies that are supported by learning technologies. A virtual university may be an institution that uses the information and communication technologies for its core activities such as providing learning opportunities, administration, materials development and distribution, delivering teaching and tuition, and providing counseling, advising and examinations. On the other hand, a virtual university may also be a virtual organization created through partnerships between traditional universities and other educational institutes. In addition, the traditional campus universities may be regarded as virtual universities if they offer learning opportunities via the Internet or combine traditional ways of learning with e-learning (Ryan et al., 2000). E-learning sets new requirements for universities: they have to build global learning infrastructures, course material has to be offered also in digital form, course material have to be distributed via the Internet and learners must have access to various virtual universities. A problem is that the current virtual university portals provide heterogeneous functionalities, which in turn hampers the learner in accessing various virtual universities. The main goal of the ONES-project is to investigate the ways of integrating various virtual universities in a way that such an aggregated virtual university would be as easily accessible for a learner as a single virtual university. Achieving such a goal requires mutual understanding of the used technology and standardized descriptions of the learning objects. Furthermore, searching

0

from various virtual universities requires mutual understanding of the information retrieval model to be used. We argue that keywords-based search (i.e., the Boolean model), though well suited for general Web searches, is unsuitable for the virtual universities’ purposes. Instead, the vector model (on which our implemented search engine is also based on) seems to be more appropriate as it provides a similarity measure (i.e., the learning object having the best match is presented first. We also introduced two interpretations for the hierarchical ontologies, which allow increasing the power of the used metadata descriptions. And finally, we also compare the performance of four algorithms for computing the similarities of the profiles. It turned out that our developed Fuzzy Matching algorithm requires less computing time as the other “exact matching” algorithms represented in the literature.

rEFErEncEs Baeza-Yates, R., & Ribeiro-Neto, B. (1999). Modern information retrieval. New York: Addison Wesley. Bentley, J., Weide, B., & Yao, A. (1980). Optimal expected-time algorithms for closest point problem. ACM Transactions on Mathematical Software, 6(4), 563-580. Fensel, D. (2001). Ontologies: Silver bullet for knowledge management and electronic commerce. Berlin: Springer Verlag. Fridman, N. N., & McGuinness, D. L. (2001, March). Ontology development 101: A guide to creating your first ontology (Stanford Knowledge Systems Laboratory Technical Report KSL-01-05, Stanford Medical Informatics Technical Report SMI-2001-0880). Friedman, J., Bentley, J., & Finkel, R. (1977). An algorithm for finding best matches in logarithmic

Information Retrieval in Virtual Universities

expected time. ACM Transactions on Mathematical Software, 3(3), 209-226. Garcia-Molina, H., Ullman, J., & Widom, J. (2000). Database system implementation. New Jersey: Prentice Hall. Gilliland-Swetland, A. J. (2000). Introduction to metadata, setting the stage. Retrieved December 20, 2004, from http://www.getty.edu/research/institute/standards/intrometadata/ Gruber, T. R. (1993, March). Toward principles for the design of ontologies used for knowledge sharing. In Padua Workshop on Formal Ontology (p. 23). Holzinger, A., Kleinberger, T., & Müller, P. (2001). Multimedia learning systems based on IEEE Learning Object Metadata (LOM). In Proceedings of ED-MEDIA 2001, Tampere, Finland. Jokela, S. (2001). Metadata enhanced content management in media companies. In Acta Polytecnica Scandinavica. Mathematics and computing series no. 114. Doctoral thesis, Helsinki University of Technology. Kassanke, S., El-Saddik, A., & Steinacker, A. (2001). Learning objects metadata and tools in the area of operations research. In Proceedings of ED-MEDIA 2001, Tampere, Finland. Lamminaho, V. (2000). Metadata specification: Forms, menus for description of courses and all other objects. CUBER project, Deliverable D3.1. Liu, J., Chan, S., Hung, A., & Lee, R. (2002). Facilitators and inhibitors of e-learning. In L. C. Jain, R. J. Howlett, N. S. Ichalkaranje, & G. Tonfoni (Eds.), Virtual environments for teaching and learning, Series on innovative intelligence (Vol. 1, pp. 75-109). World Scientific. Niemi, H. (2002). Empowering learners in the virtual university. In H. Niemi, & P. Ruohotie (Eds.), Theoretical understandings for learning

in the virtual university. University of Tampere, Research Center for Vocational Education and Training. Pöyry, P., Pelto-Aho, K., & Puustjärvi, J. (2002). The role of meta data in the CUBER system. In Proceedings of the Annual Conference of the SAICSIT 2002 (pp. 172-178). Pöyry, P., & Puustjärvi, J. (2003). CUBER: A personalised curriculum builder. In Proceedings of the 3rd IEEE International Conference on Advanced Learning Technologies, Athens, Greece (pp. 326-327). Ryan, S., Scott, B., Freeman, H., & Patel, D. (2000). The virtual university. The Internet and resourcebased learning. London: Kogan Page. Stojanovic, L., Staab, S., & Studer, R. (2001). E-learning based on the Semantic Web. In Proceedings of WebNet2001 — World Conference on the WWW and Internet, Orlando, FL. Strijker, A. (2001). Using metadata for re-using material and providing user support tools. In Proceedings of ED-MEDIA 2001, Tampere, Finland. Teare, R., Davies, D., & Sandelands, E. (1999). The virtual university — An action paradigm and process for workplace learning. Cassell. Turpeinen, M. (2000). Customizing news content for individuals and communities. In Acta Polytechnica Scandinavica. Mathematics and computing series no. 103. Doctoral thesis, Helsinki University of Technology. Vasudevan, V. (2001). A Web service primer. Retrieved December 20, 2004, from http://www. xml/lpt/a/2001/04/04/Webservices/indeax.html Yan, T., & Garcia-Molina, H. (1994). Index structures for selective dissemination of information under the Boolean Model. ACM Transactions on Database Systems, 19(2), 332-364.

0

Information Retrieval in Virtual Universities

Yli-Koivisto, J., & Puustjärvi, J. (2002). CoMet: An electronic newspaper prototype. Workshop on XML in Digital Media. In Proceedings of the 8th International Conference on Distributed Multimedia Systems (DMS’2002) (pp. 703-707).

This work was previously published in International Journal of Distance Education Technologies, Vol. 3, No. 4, edited by S.K. Chang & T.K. Shih, pp. 6-18, copyright 2005 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).

0

0

Chapter XIV

A Web-Based Tutor for Java™: Evidence of Meaningful Learning Henry H. Emurian University of Maryland, Baltimore County, USA

absTracT Students in a graduate class and an undergraduate class in Information Systems completed a Web-based programmed instruction tutor that taught a simple Java applet as the first technical training exercise in a computer programming course. The tutor is a competency-based instructional system for individualized distance learning. When a student completes the tutor, the student has achieved a targeted level of understanding the code and has written the code correctly from memory. Before and after using the tutor in the present study, students completed a software self-efficacy questionnaire and a test of the application of general Java principles (far transfer of learning). After completing the tutor, students in both classes showed increases in software self-efficacy and in correct answers on the test of general principles. These findings contribute to the stream of formative evaluations of the tutoring system. They show the capacity of the Web-based tutor to generate meaningful learning (i.e., understanding of concepts) at the level of the individual student.

inTroDUcTion This chapter presents a continuation of a series of formative evaluations to assess and to enhance the instructional effectiveness of an automated and individualized distance learning system that is intended to assist Information Systems students in beginning their study of Java™. We previously reported our progress in the development of this tutoring system, which teaches a simple

Java applet, and its application as the first technical training exercise for students in a computer programming course (Emurian, 2004; Emurian & Durham, 2001, 2002; Emurian, Hu, Wang, & Durham, 2000). The purpose of the tutor is to provide each and every student with a documented and identical level of elementary knowledge and skill. The tutoring system has been demonstrably effective in promoting technical skill and in cultivating self-confidence in beginning students

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

by giving them a successful learning experience that motivates their further study of Java using textbooks, lectures, laboratory demonstrations, independent problem solving, and the like. The tutoring system is targeted to Information Systems majors, whose primary interests may lie in systems analysis and design, database, decision support systems, knowledge management, and information resource management, but who would benefit from acquiring elementary skill in an object-oriented programming language such as Java. One of the challenges of developing an automated distance learning system, however, is to craft the instructional experience so that students acquire the capability to solve problems not explicitly taught or encountered in the system itself. When students are able to apply knowledge successfully to new situations, they are said to be demonstrating meaningful learning (Mayer, 2002) as opposed to reciting facts acquired by rote memorization. These two outcomes reflect the opposite endpoints on a generality-specificity dimension of skill (Novick, 1990). Generalizable rules, which may be the essence of meaningful learning, can be acquired by direct instruction and rehearsal or by induction, when many different situations are encountered that exhibit the general rule (Kudadjie-Gyamfi & Rachlin, 2002). The former tactic is consistent with our instructional system design, which is competency-based and intended to insure that all students reach the same level of knowledge and skill with the applet being taught. The theory supporting the development of the tutoring system is a behavior-analytic model based upon the learn unit formulation of Greer and McDonough (1999) as applied to programmed instruction for technology education (Greer, 2002). The interactive tutoring system interfaces that are presented to the learner reflect the systematic increase in the size of the learn units from simple symbol production (i.e., learning to type) to writing and understanding the entire program. The

0

stream of work leading up to the present study constitutes a series of replications over which the system was refined to ensure that mastery would occur for all learners. This approach is similar to instructional material improvements suggested by formative evaluation (Harley, Seals, & Rosson, 1998) and by design-based research (Hoadley, 2004), and how this strategy contrasts with between-subjects evaluations is discussed in subsequent paragraphs. The purpose of the present study is to show that students who complete the tutor do acquire general rules that are applicable to programming problems not explicitly addressed in the tutor itself. The study extends our prior investigation (Emurian, 2005) by increasing the number of testable rules to 10 and by supporting the reliability of the outcomes over two different groups of students. This research approach constitutes systematic replication (Sidman, 1960), which is an alternative to null hypothesis testing and intended to validate externally this particular instructional system design rather than to test hypotheses regarding effect sizes across alternative designs. This methodology falls within the scope of an outcomes assessment model of evaluating teaching effectiveness (Fox & Hackerman, 2003), and, in the present situation, the teacher is the Web-based tutoring system. Interpretative surveys of the scientific literature in far transfer effects of learning continue to show the advantages of explicitly teaching generalizable principles and rules rather than expecting such knowledge to develop implicitly or abstractly as a byproduct of memorizing facts (Barnett & Ceci, 2002) or by pure discovery learning (Mayer, 2004). For example, it is likely more efficient to teach students the rule to begin the name of a Java class with a capital letter than to expect students to discover such a rule inductively by studying many different programs and by trying to discern commonalities among them. In fact, a combination of teaching rules with examples might be optimal for meaningful learning, and

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

our approach to the design of the tutoring system is based on that latter assumption. The study to follow, then, assesses the extent of rule-governed knowledge before and after students have used the tutoring system.

mEThoD subjects Two groups of students participated as subjects in this investigation. Prior to using the tutor, the subjects completed a questionnaire that collected information on gender, age, number of prior programming courses taken, and experience with Java. The anchors for the latter rating scale were as follows: 1 = No experience. I am a novice in Java. to 10 = Extensive experience. I am an expert in Java. The subjects in Group 1 were eight female and four male graduate students in Information Systems, who were enrolled in a course (Summer 2003) entitled Graphical User

Interface Systems Using Java. The subjects in Group 2 were two female and 10 male undergraduate students enrolled in the same course at a later time (Spring 2004). For Group 1, the median age was 27 years (range = 21 to 49), and for Group 2, the median age was 22 years (range = 20 to 26). A Kruskal-Wallis test1 showed a significant difference between the two groups in age (χ2 = 5.68, df = 1, p < 0.02). Figure 1(a) presents boxplots of the number of programming courses that the students had previously taken and their experience with Java2. The median number of courses taken was three for both groups, although the interquartile range was graphically higher for Group 2. The median reported Java experience was two for both groups. Differences between the groups were not significant on either measure.

materials The tutoring system teaches a simple Java applet, which is a program that is downloaded from a server and run in a browser. The program, pre-

Figure 1. (a) Boxplots of the number of programming courses that the students had previously taken and their experience with Java. (b) Boxplots of overall evaluation, teaching effectiveness, and usability ratings. Circles are outliers, and triangles are extreme values.

0

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

sented in Appendix A, displays a text string in a browser window. The tutoring system is freely available on the Web3. It is intended for students who are not proficient computer programmers and who may lack confidence in their ability to write and to understand a program that works. The Applet code is organized into 32 items and 10 lines (i.e., rows). The student is taught the meaning of each item and the meaning of each line. The student must pass a multiple-choice test on each program component; studying continues until each test is passed correctly. The tutor exits when the student can write the entire program without an error. The content of the tutor, to include the tests embedded within the tutor interfaces, is freely available as a document 4. The seven stages of the tutoring system, from basic instructions and code examples to the construction of the code from memory, are presented in Emurian (2004) and Emurian, Wang, and Durham (2003). In brief, the tutoring system is an interactive system that combines teaching, assessment of competency, and rehearsal within a single framework. The design of the system is based upon programmed instruction, which takes a learner through a series of experiences from simple mastery of the form of symbols to writing and understanding a complete program. This design reflects the application of behavior principles to designing teaching strategies for technology education (Greer, 2002). These principles are uniquely applicable to individualized instruction. Information is delivered to the student in a frame. A frame consists of (1) the presentation of the Java symbol (e.g., import) to be learned, within the proper context; (2) a textual display of information explaining the symbol’s meaning and use; (3) a multiple-choice test on the meaning of the symbol; and (4) an input field for typing the symbol by recall. If the student makes an error during steps 3 or 4, the tutor resets to step 1, and that cycle repeats until the input is correct. When the input in step 4 is correct, the student

0

progresses to the next frame. The functional properties of a frame constitute a learn unit (Greer & McDonough, 1999) when the student progresses from one frame to the next, determined by accurate performance. Next is presented the explanation for the ninth of the 32 item frames in the tutoring system. It explains the meaning of the MyProgram item, which is a subclass of the Applet class in the program.

myProgram Explanation The term MyProgram is the name of the class that you are writing. Your Java program is a class. The name is an arbitrary alphanumeric string. MyProgram is not the name of an instance of this class. It is the name of the class. It is important that you begin to distinguish the name of a class from the names of particular instances of that class that are created later. This distinction will become clearer as you progress through the tutor. Notice that the name of the class begins with a capital letter. That is a convention in Java. The name of a class begins with a capital letter. That is an important rule to know. The text file that contains the Java program for the MyProgram class must have exactly the same name, together with “dot java” at the end. The file for your program would be named MyProgram.java. The name of the text file must match exactly the name of the class. That is an important rule to know. The Java text file, which is the source program, will be compiled with javac MyProgram.java. The result of compiling the program is a class file named MyProgram.class, which will be located in your directory. The ten rule-based multiple-choice questions are presented in Appendix B. For Group 1, only the first five questions were available. For Group 2, the number of questions was increased to 10. The correct solution to each question required the application of a general principle that was

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

presented in the explanatory information. None of the items to solve these questions appeared verbatim in the explanations or in the multiplechoice tests that were embedded in the frames. This eliminated rote memorization as a reason for correct solutions, should they be observed at all. For each rule-based question, the student rated his or her confidence that the correct answer was selected. The ordinal scale anchors were 1 and 10, where 1 = Not confident and 10 = Totally confident. A 10-point rating scale was adopted to increase the sensitivity of the scale in order to detect changes in ratings over three successive assessment occasions. Figure 2 presents the running applet as it appears within the Mozilla browser. Figure 3 presents the display of the 12th Java symbol (i.e., item) to be learned, which is an opening brace symbol ({). In Figure 3, the { would appear when the learner selects the enabled Show Java button. In the browser, the symbol is blue to set it apart from the previous symbols that have been learned. Figure 4 shows the symbol’s explanatory information that is presented when the user selects

the enabled Explain button. Figure 5 shows the multiple-choice test for this particular opening brace in the program. Figure 6 shows the input field for the brace. Figure 7 shows the interface for the line-by-line input. In this example, the code displayed in the third line (i.e., Row 3) is ready to be assessed when the learner presses the keyboard Enter key. This interface functions similarly to the items learning interface, but the unit of learning is a line of code, and there are two iterations through the interface. The second iteration has no tests associated with it. Figure 8 shows the final interface in the tutor, where the learner enters the entire program. The input format previously enforced by the tutor is relaxed for this stage of learning compared to the previous tutor stages, and the input is evaluated as a stream of characters. The remaining instructional component of the tutoring system, along with the instructions to the learner, may be observed by running the tutor on the Web.

Figure 2. The applet as it appears running in a browser on the Web



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Figure 3. The display of the Java symbol to be learned, which is the opening brace in this example. The brace, as are all symbols first observed for learning, is colored blue to distinguish it from the symbols that already have been learned in the sequence of presentations. The symbol to be learned is presented when the user selects the enabled Show Java button.

Figure 4. The explanatory information for the brace. It is presented when the user selects the enabled Explain button.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Figure 5. The multiple-choice test for this particular opening brace in the program. It is presented when the user selects the enabled Test button.

Figure 6. The input field for the symbol. The empty field is presented at the location of the symbol in the tutor code, and its width matches the width of the characters to be entered. This view is presented automatically when the user selects the correct test choice. The user types the symbol from memory into the input field and then hits the Enter key on the keyboard. If the input is correct, the symbol is locked into position in the display, and it appears colored black.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Figure 7. An example of the line-by-line interface with input in the third row. This interface is functionally similar to the items learning interface, but the unit of learning is a line of code, not an item of code.

Figure 8. The final stage in the tutor in which the learner enters the entire program. If the input contains an error, the correct code is displayed. After the code view is closed, the learner again attempts to enter the program into a cleared input field. This cycle repeats until the program is entered correctly.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Procedure During the first scheduled class, students used the Web-based tutor. Although the students accessed the tutor during a class located in a PC lab, the system is designed as an individualized distance learning system. All students completed all stages in the tutor within the three-hour class period. This means that all students in both groups left the first class period being able to write the Java code from memory and with no error. The students had also studied the frames until they accurately could input all items and lines and correctly answer all multiple-choice test questions. Prior to using the tutor, students completed a questionnaire that assessed software self-efficacy. For each of the 21 unique items of code in the program, the student rated his or her confidence in being able to use the symbol, where 1 = Not confident and 10 = Totally confident. This was the measure of software self-efficacy, based on the original work by Bandura (1977) and the subsequent adoption of this approach by researchers in information technology education (Potosky, 2002; Torkzadeh & Van Dyke, 2002). Students also completed the five (Group 1) or 10 (Group 2) rule-based questions, to include the rating of confidence in the accuracy of the answer selected. After the students completed the tutor, they repeated the software self-efficacy ratings and rule-based questions. The students then rated the overall quality of the tutor, the effectiveness of the tutor in learning Java, and the usability of the tutor interfaces. Each of these scales was a 10-point ordinal scale, where 1 = Poor quality and 10 = Best quality. During the second class period, the author repeated the teaching of the Applet, but this time, a lecture and discussion format was used. This was part of the formative evaluation, and insights gained here are applicable to refining the tutoring system. For Group 1, the second class was two days after the first one. For Group 2, the second class

was one week after the first one. These different delays were attributable to the days designated for the classes to meet. During the second class, the author wrote the program on the board and discussed each item and line of code. The students simultaneously entered the program into a UNIX™ text editor. At the completion of the lecture, the students were taught how to compile the program into bytecode. Additionally, the UNIX directory tree and file protections were taught. The HTML file then was taught, and the students ran the Applet on the Web. The students then repeated the questionnaire assessing software self-efficacy and rule-based learning. All of this instructional material, however, was also presented on the Web as part of the ancillary material supporting the individualized tutor.

rEsUlTs Figure 9 presents boxplots of total correct rulebased answers by all students across the three assessment occasions: pre-tutor, post-tutor, and post-lecture. The figure shows graphically that the median value increased over the three occasions, and a Friedman test was significant for Group 1 (χ2 = 19.58, df = 2, p < 0.001) and for Group 2 (χ2 = 10.21, df = 2, p < 0.001). The figure also shows that the most pronounced increase occurred between the pre-tutor and post-tutor occasions, in comparison to the post-tutor and post-lecture occasions. For Group 1, a comparison of the means of the differences for all 12 students between pre-tutor and post-tutor totals (Mean = 2.3) with post-tutor and post-lecture totals (Mean = 0.3) was significant t(17.5) for unequal variances = 4.24, p = 0.001. For Group 2, a comparison of the means of the differences for all 12 students between pre-tutor and post-tutor totals (Mean = 1.5) with post-tutor and post-lecture totals (Mean = 0.7) was not significant, t(16.9) for unequal variances = 1.12, p > 0.10).



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Figure 9. Boxplots of total correct rule-based answers by all students across the three assessment occasions: pre-tutor, post-tutor, and post-lecture. The triangle is an extreme value.

Figure 10. Boxplots of median confidence ratings for Right and Wrong answers for pre-tutor (Pre), posttutor (Post), and post-lecture (Lecture) assessment occasions. Data for the Lecture occasion were not obtained for one student in Group 1. Circles are outliers, and triangles are extreme values.

Figure 10 presents boxplots of median confidence ratings for Right (R) and Wrong (W) answers for pre-tutor (Pre), post-tutor (Post), and post-lecture (Lecture) assessment occasions for Group 1 and Group 2. Values in the boxplots are based upon the collection of median ratings for each student for R and W answers across the three occasions. For Group 1, one student did



not report data for the post-lecture assessment, and n = 11. For Group 1, K-W comparisons between R and W ratings were not significant within pre-tutor (χ2 = 1.07, df = 1, p > 0.10) and post-tutor occasions (χ2 = 0.16, df = 1, p > 0.10). Accordingly, data were combined for R and W within each occasion, and a K-W comparison between pre-tutor and post-

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

tutor ratings was significant (χ2 = 19.68, df = 1, p < 0.001). Too few medians were present in the W category for the post-lecture assessment for a meaningful comparison using those data. For Group 2, K-W comparisons between R and W ratings were not significant within pre-tutor (χ2 = 0.62, df = 1, p > 0.10) and post-tutor occasions (χ2 = 0.11, df = 1, p > 0.10). Accordingly, data were combined for R and W within each occasion, and a K-W comparison between pre-tutor and post-tutor ratings was significant (χ2 = 23.27, df = 1, p < 0.001). A comparison with the postlecture outcomes was not undertaken because a rating ceiling (i.e., median = 10) was graphically apparent on the post-tutor occasion. These ratings show the students’ insensitivity in monitoring their own learning. Since the null hypothesis of no difference in confidence ratings between R and W answers could not be rejected for Group 1 and Group 2, the learners did not know, perhaps, that their learning was incomplete. Since self-regulation of learning is an important skill (Veenman, Prins, & Elshout, 2002; Young, 1996; Zimmerman, 1994), how to achieve this outcome within the context of the present tutoring system

warrants consideration as this teaching technology continues to mature. Figure 11 presents boxplots of self-reports of software self-efficacy by all students across the three assessment occasions: pre-tutor (Cronbach’s alpha = 0.97 for Group 1 and 0.97 for Group 2, post-tutor (Cronbach’s alpha = 0.98 for Group 1 and 0.98 for Group 2), and post-lecture (Cronbach’s alpha = 0.98 for Group 1 and 0.98 for Group 2). For Group 1, the figure shows graphically that the median value increased over the three occasions, and a Friedman test was significant (χ2 = 21.38, df = 2, p < 0.001). The figure also shows that the most pronounced increase occurred between the pre-tutor and post-tutor occasions compared to the post-tutor and post-lecture occasions. A comparison of the means of the differences for all 12 students between pre-tutor and post-tutor ratings (Mean = 5.2) with post-tutor and post-lecture ratings (Mean = 1.0) was significant, t(17.3) for unequal variances = 4.80, p < 0.001. For Group 2, Figure 11 shows graphically that the median value increased between the pre-tutor and post-tutor occasions, and it reached the ceiling on that latter occasion. A Friedman test

Figure 11. Boxplots of self-reports of software self-efficacy by all students across the three assessment occasions. Circles are outliers, and triangles are extreme values.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Figure 12. Boxplots of errors on the program interface requiring the input of the entire Java program. A single error reflected any and all errors that were present when the code was submitted for evaluation. An error, then, reflected an incorrect attempt to write the code correctly. The data records for two students in Group 1 and one student in Group 2 were not usable. Circles are outliers, and triangles are extreme values.

was significant (χ2 = 15.44, df = 2, p < 0.001). A comparison of the means of the differences, for all 12 subjects between pre-tutor and post-tutor ratings (Mean = 4.2) with post-tutor and post-lecture ratings (Mean = 0.1) was significant, t(11.1) for unequal variances = 3.96, p = 0.002. Figure 12 presents boxplots of errors on the final tutor interface, the program interface, which required writing the Java code (see Figure 8). When the student submitted the code for evaluation, one or more errors anywhere in the code produced a pop-up window displaying the correct code. It was that event that was counted as an error. When that window was closed by the student, the program interface was cleared, and the student again entered the code from memory. The tutor is designed as a series of applets, and the performance data are generated and saved automatically for each student. The data for two students in Group 1 and one student in Group 2



were corrupted, thereby reducing the number of students to 10 and 11, respectively, for this analysis. Figure 12 shows that the median errors for students in Group 1 was higher than for Group 2, and a K-W test showed that the difference between the groups was marginally significant (χ2 = 2.82, df = 1, p < 0.10). Figure 1(b) shows post-tutor ratings on the Overall, Learning Effectiveness, and Usability scales. For both groups, all medians were nine or higher, with Group 1 showing a median of 10 for the Overall rating. A K-W test showed that the difference between Group 1 and Group 2 ratings on the Overall scale was significant (χ2 = 4.67, df = 1, p < 0.05). The extreme value (i.e., rating = 1) present for the Overall and Learning Effectiveness scales was reported by a computer science major who had a different history of computer programming compared to the other students. The tutor is best received by students with minimal

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

background in programming, and more advanced students, even those with no experience in Java, perhaps should be excused from using such a tutor as the one presented here. Otherwise, the data show that the tutor is well received by graduate and undergraduate students who are taking this course.

DiscUssion The results of this study provide further evidence of the effectiveness of programmed instruction presented as an individualized distance learning system in technology education. Students with varying backgrounds all achieved a common outcome of mastering a simple Java applet as the first technical exercise in a programming course. Students not only achieved tested mastery on the syntax and semantics of the code and the capability to recite the code, but they also demonstrated meaningful learning as evidenced by improvements on rule-based questions that required far transfer knowledge. Although studying and memorizing code may well be valuable to achieving facility using a new set of symbols, at the very least the evidence here, which indicates that frames of information and corresponding tests of understanding and retention may be crafted to promote the acquisition of problem-solving skills, supports the more general usefulness of the programmed instruction tutoring system in promoting cognitive skill development. The present study, then, extends the generality of our previous investigation (Emurian, 2005) to a situation that involved more rule questions with two new groups of students. This shows the reliability of the outcomes over systematic replications under actual classroom conditions. The merits of this design-based research tactic for studying learning in context, which involves students in a classroom, have been discussed recently in a special issue of Educational Psychologist (Special Issue, 2004).

The results are consistent with recommendations for evaluating the effectiveness of science, technology, engineering, and mathematics (STEM) instruction: “Provide experiences for students to develop functional understanding. These [exemplary] programs place emphasis on students’ understanding of science concepts and ability to apply these concepts to new situations” (McCray, DeHaan & Schuck, 2003, p. 30). Having demonstrated the far transfer consequences of using the tutoring system as an experience that fosters understanding, research in this area can focus on the refinement of the instructional information based upon a relational-frame theory of cognition (Hayes, Fox, Gifford, Wilson, BarnesHolmes & Healy, 2001) that might be anticipated to potentiate such meaningful learning outcomes of programmed instruction. For example, answering rule-question number eight correctly required a combinatorial entailment of the information presented in the corresponding explanation in the tutor (Hayes et al., 2001). How to optimize such frames to enhance functional understanding is an important consideration for future research in this area of individualized instructional design applicable to distance learning technologies. In that regard, designers of prose for textbook learning recommend the use of signaling, adjunct questions, and advance organizers as techniques for enhancing a learner’s understanding of the material (Mayer, 2002). Unfortunately, perhaps, the research leading to such a recommendation typically is based only on single and time-limited episodes of studying across various experimental treatments. That strategy reveals little about the value of actual studying behavior and resulting understanding, because the process of studying until mastery has occurred is one hallmark of a high-achieving, self-regulated student (Eilam & Aharon, 2003; Pintrich, 2003). How to make a learning process leading to mastery available to all students should be the goal of instructional design. Equally unfortu-



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

nate, moreover, is the entrenched elitism of our educational system that continues to undervalue teaching effectiveness. At present, faculty members are likely to face significant disincentives to learn new teaching approaches and reformulate an introductory course: it requires a large investment of time, it is a distraction from the focus on research, and their investment may not be rewarded. (McCray, DeHaan & Schuck, 2003, p. 51) There, then, are barriers to implementing effective instruction that is directed toward the achievement of the individual learner. The importance of the ratings of software self-efficacy is to be understood in terms of the impact of the learning experience on the students’ motivation to continue their studies. Given the frequently expressed concern that women and minority groups avoid or withdraw from STEM disciplines (Emurian, 2004), it is encouraging to observe that at least some instructional tactics, such as programmed instruction, may be helpful to provide both skill and confidence in students who initially are interested in a discipline, but whose lack of preparation may sometimes result in demoralization to the point of avoiding or withdrawing from continued study. The essence of effective automated tutoring is to provide a set of experiences that gives all students the skill and confidence to manage their own learning effectively without regard to content and without the continued support of a tutoring system. The ratings of confidence in the rule-based answers, however, present a more complex interpretative challenge. There was insufficient evidence to suggest that students within either group were able to distinguish between answers that were correct and answers that were incorrect, at least based upon the students’ confidence ratings. This was evident, moreover, even within the context of a general improvement in performance between pre-tutor and post-tutor occasions.

0

As noted previously, self-regulation of learning (learning to learn) is an important skill, and these data suggest that without feedback for correct and incorrect answers, learners may not be able to recognize the extent of their own competencies and deficiencies. Although an experimental analysis of rule-governed performance may require obtaining information under the present set of conditions, a classroom application requires feedback (i.e., knowledge of results) for a student’s performance errors, as was the case with the assessment tools that were embedded within the tutor itself (Locke, Chah, Harrison, & Lustgarten, 1989; Locke & Lantham, 2002). As indicated by McCray et al. (2003), STEM students need to know when they do not understand, and they need the experiences that lead to understanding a body of material at a deep level. These confidence ratings, then, may reveal a shortcoming in the tutor design that was not evident by a simple tally of errors. In our series of studies, we make an attempt to document individual differences among subjects and groups only to show the generality of the tutoring system to have a positive impact on learners with different backgrounds, preparatory experiences, and evaluations of the tutor. In the present study, for example, the graduate students as a group were somewhat older than the undergraduates, but on measures of previous courses taken and Java experience, the undergraduates and graduates were not found to differ. Furthermore, there were more female students in the graduate course than in the undergraduate course. The number of errors made on the program interface was larger for the graduate students than for the undergraduates, and the graduate students gave higher overall ratings to the tutoring system compared to the undergraduates. Despite these differences, all students exited the tutoring system with the same level of knowledge and skill, at least with respect to the targeted learning objectives, because the system was designed to achieve that outcome by individualized distance learning. Had the system been designed to require the understanding of

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

the concepts tested by the rule-based questions that were administered, all students would have exited the tutor with a history of answering all rule-based questions accurately. Fostering constructive metacognitive processes and supporting individual differences in ability constitute the foundation of effective automated tutoring (Cuevas, Fiore, Bowers & Salas, 2004). In addition to programmed instruction, technology education also would benefit from a consideration of the use of direct instruction (Watkins & Slocum, 2004), precision teaching (Chiesa & Robertson, 2000), and interteaching (Boyce & Hineline, 2002), as those tactics facilitate task mastery at the level of the individual learner. Moreover, the learning history generated by the programmed instruction tutoring system is intended to free our students from future dependence on such automated instruction. With the ultimate goal of producing self-directed learners, our overall strategy in the course is gradually to withdraw structured support over the semester while teaching students to seek out information that is required to solve increasingly more difficult programming problems. Other teaching tactics, such as the Online Problem-Based Model of Learning Java (Tsang & Chan, 2004), have an important role to play in providing subsequent experiences that promote such self-directed learning. Our entire course, not just the Java tutor, is freely available on the Web5, and our approach can be observed, experienced, and adopted by others. As discussed elsewhere (Emurian & Durham, 2003), much of the literature in teaching computer programming addresses this challenge as though the skill of computer programming requires a unique teaching technology. The research also is directed toward groups of students rather than to the individual learner. Our approach is different. We assume that learning to write computer programs falls within the scope of training in general (Salas & Cannon-Bowers, 2001) and rulegoverned learning in particular (Hayes, 1989). We also assume that a teaching technology only

can be rationally developed and applied when it is directed toward the achievement of a criterion of mastery by each and every student. In that latter regard, research and interventions that are based on null hypothesis refutations of average performance differences between and among treatment groups by definition accept at least some deficient student performance as an outcome. It is encouraging, then, to see the emergence of more achievement-oriented research based on pre-training and post-training comparisons in one group of learners (Potosky, 2002; Torkzadeh & Van Dyke, 2002) in contrast to between-subjects comparisons. As indicated by Sackett and Mullen (1993), it may be more important to an organization to know that an instructional intervention will be successful for all learners than it is to know that average performances between and among groups show differences that are statistically significant. Null hypothesis research typically is focused on effect size differences across treatment conditions, and that approach has recently been characterized by one education scholar as an “old fashioned experimental ‘horse-race’ design” (Mayer, 2004, p. 16). Although some proponents of education research continue to advocate that latter model of randomized designs (Towne & Hilton, 2004), how a shift away from null hypothesis evaluation toward alternative methodologies might be undertaken has been discussed elsewhere (Emurian, 2005). In fact, the process of systematic replication used here (Sidman, 1960) is consistent with Design Principle 5 presented in the Shavelson and Towne (2002) report: “Replication and generalization strengthen and clarify the limits of scientific conjectures and theories” (p. 70). Once a behavioral teaching strategy has been demonstrated to be effective for all students, experimental evaluations then might be undertaken to determine the optimal parameters of the instructional system, but only after the initial design has been demonstrated to be effective at the level of the individual learner.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

It is an unfortunate irony, however, that educational research sometimes is perceived as less valuable than other areas of research, at least within the social sciences (Sternberg & Lyon, 2002). The irony comes from the obvious importance of education and of knowing and applying the conditions that will help students to learn best throughout the life span. In that regard, research in how best to teach computer programming typically is characterized by comparing average performance among several groups of learners, where each group is exposed to a somewhat different instructional condition. The literature is filled with scores of such studies ranging from early evaluations of conditional constructions (Sime, Green, & Guest, 1973) through LOGO programming (Pea & Kurland, 1984) to levels of graphical support in teaching UNIX™(Sohn & Doane, 1997), classroom tactics for teaching computer programming (Mayer, 1988), and structured exercises and guided exploration in learning the Hypercard™ program (Wiedenbeck & Zila, 1997). That strategy inherently is flawed, because it tacitly accepts the outcome that not all students will achieve mastery even in the group with the best average performance. When it comes to learning Java, the focus on the individual learner is acknowledged by textbook authors such as Deitel and Deitel (1999): “All learners initially learn how to program by mimicking what other programmers have done before them” (p. 38). As distance-learning technologies continue to mature, so will the requirement for the design of individualized instruction rather than grouporiented instruction suitable for implementation with such technologies. The motivation for the strategy of betweensubjects comparisons, of course, is to determine the best instructional tactic, but the outcomes of the research are useful practically only when the time and resources for instructional delivery or for studying are constrained for all students. Such constraints have nothing to do with the process of learning. To adopt the best teaching



strategy should mean to respect the right of all students to be given the opportunity to achieve mastery, where opportunity is redefined to mean sustained exposure to the proper conditions of learning to include being taught learning strategies (Namlu, 2003) until achievement has been attained at the level of the individual student. Programmed instruction is a promising tool to foster technical mastery and meaningful learning for students who would benefit from a set of principled interactive instructions that assures the achievement of competency, confidence, and meaningful learning. The limitations of this work are attributable to it being a series of formative evaluations, rather than an experiment in the traditional sense, although we have argued against such a traditional research methodology. Our interest is effect efficiency and dependability for a common learning outcome for all students rather than effect size differences among alternative instructional treatments producing variable learning outcomes across students. In that regard, the components and parameters of the tutoring system have not been independently validated or isolated for effectiveness. For example, it is not obvious that all learners benefit from the introductory interfaces that only provide facility with the symbols, independent of their meaning. It also is not obvious how much repetition should be built into the tutor. In its current version, for example, the tutor requires two iterations through the line-by-line interface and only one correct input of the entire applet program. Data supporting those parameters are lacking. The choice was made heuristically by our efforts to make the tutor fit into a threehour period for most students. Since overlearning and repetition are important to retention of a skill (Durham & Emurian, 1998; Swezey & Llaneras, 1997), determining the optimal set of parameters will require further experimental evaluation. Can students achieve the same outcome with other instructional approaches? Almost certainly, some can. Can students simply study a manual

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

or textbook and come away with equivalent understanding and skill? Quite possibly, many students can. However, when a teacher wants to be certain that all students achieve a targeted level of performance with a Java applet, the Webbased programmed instruction tutoring system is dependable in producing that outcome, and it is generally well received by students. The individualized tutor will continue to play an important role in our offerings to Information Systems majors, whether as a component in a face-to-face course or within our online distance-education programs.

rEFErEncEs Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128, 612-637. Boyce, T. E., & Hineline, P. N. (2002). Interteaching: A strategy for enhancing the user-friendliness of behavioral arrangements in the college classroom. The Behavior Analyst, 25, 215-226. Chiesa, M., & Robertson, A. (2000). Precision teaching and fluency training: Making math easier for pupils and teachers. Educational Psychology in Practice, 16(3), 297-310. Cuevas, H. M., Fiore, S. M., Bowers, C. A., & Salas, E. (2004). Fostering constructive cognitive and metacognitive activity in computer-based complex task training environments. Computers in Human Behavior, 20(2), 225-241. Deitel, H. M., & Deitel, P. J. (1999). Java™ How to program (3rd ed.). Upper Saddle River, NJ: Prentice Hall.

Durham, A. G., & Emurian, H. H. (1998). Learning and retention with a menu and a command line interface. Computers in Human Behavior, 14, 597-620. Eilam, B., & Aharon, I. (2003). Student’s planning in the process of self-regulated learning. Contemporary Educational Psychology, 28(3), 304-334. Emurian, H. H. (2004). A programmed instruction tutoring system for Java™: Consideration of learning performance and software self-efficacy. Computers in Human Behavior, 20(3), 423-459. Emurian, H. H. (2005). Web-based programmed instruction: Evidence of rule-governed learning. Computers in Human Behavior, 21, 893-915. Emurian, H. H., & Durham, A. G. (2001, May 20-23). A personalized system of instruction for teaching Java. In M. Khosrow-Pour (Ed.), Managing information Technology in a Global Economy, 2001 Information Resources Management Association International Conference (pp. 155-160), Toronto, Ontario, Canada. Hershey, PA: Idea Group Publishing. Emurian, H. H., & Durham, A. G. (2002, May 19-22). Enhanced learning on a programmed instruction tutoring system for JAVA. In M. Khosrow-Pour (Ed.), Issues & Trends of Information Technology Management in Contemporary Organizations, 2002 Information Resources Management Association International Conference (pp. 205-208), Seattle, Washington, USA. Hershey, PA: Idea Group Publishing. Emurian, H. H., & Durham, A. G. (2003). Computer-based tutoring systems: A behavioral approach. In J. A. Jacko, & A. Sears (Eds.), Handbook of human-computer interaction (pp. 677-697). Mahwah, NJ: Lawrence Erlbaum & Associates. Emurian, H. H., Hu, X., Wang, J., & Durham, A. G. (2000). Learning Java: A programmed



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

instruction approach using applets. Computers in Human Behavior, 16, 395-422.

in a self-control task: Effects of changes in contingencies. Behavioral Processes, 57(1), 29-35.

Emurian, H. H., Wang, J., & Durham, A. G. (2003). Analysis of learner performance on a tutoring system for Java. In T. McGill (Ed.), Current issues in IT education (pp. 46-76). Hershey, PA: IRM Press.

Locke, E. A., Chah, D., Harrison, S., & Lustgarten, N. (1989). Separating the effects of goal specificity from goal level. Organizational Behavior and Human Performance, 43, 270-287.

Fox, M. A., & Hackerman, N. (2003). Evaluating and improving undergraduate teaching in science, technology, engineering, and mathematics. Washington, DC: The National Academies Press. Greer, R. D. (2002). Designing teaching strategies: An applied behavior analysis systems approach. New York: Academic Press. Greer, R. D., & McDonough, S. H. (1999). Is the learn unit a fundamental measure of pedagogy? The Behavior Analyst, 22, 5-16. Harley, H. D., Seals, C. D., & Rosson, M. B. (1998, Winter). A formative evaluation of scenario-based tools for learning object-oriented design. ACM Crossroads, 1-5. Retrieved April 6, 2004, from http://www.acm.org/crossroads. xrds5-1/eval.html Hayes, S. C. (1989). Rule-governed behavior: Cognition, contingencies, and instructional control. New York: Plenum Press. Hayes, S. C., Fox, E., Gifford, E. V., Wilson, K. G., Barnes-Holmes, D., & Healy, O. (2001). Derived relational responding as learned behavior. In S. C. Hayes, D. Barnes-Holmes & B. Roche (Eds.), Relational frame theory: A post-skinnerian account of human language and cognition. New York: Kluwer Academic. Hoadley, C. M. (2004). Methodological alignment in design-based research. Educational Psychologist, 39(4), 203-212. Kudadjie-Gyamfi, E., & Rachlin, H. (2002). Rulegoverned versus contingency-governed behavior



Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation. American Psychologist, 57(9), 705-717. Maxwell, S. E., & Delaney, H. D. (2000). Designing experiments and analyzing data. Mahwah, NJ: Lawrence Erlbaum Associates. Mayer, R. E. (Ed.). (1988). Teaching and learning computer programming. Hillsdale, NJ: Erlbaum. Mayer, R. E. (2002). The promise of educational psychology. Volume II. Teaching for meaningful learning. Upper Saddle River, NJ: Pearson Education. Mayer, R. E. (2004). Should there be a threestrikes rule against pure discovery learning? The case for guided methods of instruction. American Psychologist, 59(1), 14-19. McCray, R. A., DeHaan, R. L., & Schuck, J. A. (Eds.). (2003). Improving undergraduate instruction in science, technology, engineering, and mathematics. Washington, DC: The National Academies Press. Namlu, A. G. (2003). The effect of learning strategy on computer anxiety. Computers in Human Behavior, 19, 565-578. Novick, L. R. (1990). Representational transfer in problem solving. Psychological Science, 1, 128-132. Pea, R. D., & Kurland, D. M. (1984). On the cognitive effects of learning computer programming. New Ideas in Psychology, 2, 137-168.

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

Pintrich, P. R. (2003). A motivational science perspective on the role of student motivation in learning and teaching contexts. Journal of Educational Psychology, 95(4), 667-686.

Swezey, R. W., & Llaneras, R. E. (1997). Models in training and instruction. In G. Salvendy (Ed.), Handbook of human factors and ergonomics (pp. 514-577). New York: Wiley.

Potosky, D. (2002). A field study of computer efficacy beliefs as an outcome of training: The role of computer playfulness, computer knowledge, and performance during training. Computers in Human Behavior, 18, 214-255.

Torkzadeh, G., & Van Dyke, T. P. (2002). Effects of training on Internet self-efficacy and computer user attitudes. Computers in Human Behavior, 18, 479-494.

Sackett, P. R., & Mullen, E. J. (1993). Beyond formal experimental design: Towards an expanded view of the training evaluation process. Personnel Psychology, 46, 613-627. Salas, E., & Cannon-Bowers, J. A. (2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471-499. Shavelson, R. J., & Towne, L. (Eds.). (2002). Scientific research in education. Washington, DC: The National Academies Press. Sidman, M. (1960). Tactics of scientific research. New York: Basic Books. Sime, M. E., Green, T. R. G., & Guest, D. J. (1973). Psychological evaluation of two conditional constructions used in computer languages. International Journal of Man-Machine Studies, 5, 105-113. Sohn, Y. W., & Doane, S. M. (1997). Cognitive constraints on computer problem-solving skills. Journal of Experimental Psychology: Applied, 3, 288-312. Special Issue. (2004). Design-based research methods for studying learning in context. Educational Psychologist, 39(4), 199-260. Sternberg, R. J., & Lyon, G. R. (2002). Making a difference to education: Will psychology pass up the chance? Monitor on Psychology, 33, 76. Retrieved April 6, 2004, from http://www.apa. org/monitor/julaug02/difference.html

Towne, L., & Hilton, M. (Eds.). (2004). Implementing randomized field trials in education. Washington, DC: The National Academies Press. Tsang, A. C. W., & Chan, N. (2004). An online problem-based model for the learning of Java. Journal of Electronic Commerce in Organizations, 2(2), 55-64. Veenman, M. V. J., Prins, F. J., & Elshout, J. J. (2002). Initial inductive learning in a complex computer simulated environment: The role of metacognitive skills and intellectual ability. Computers in Human Behavior, 18, 327-241. Watkins, C. L., & Slocum, T. A. (2004). The components of direct instruction. Journal of Direct Instruction, 3(2), 75-110. Wiedenbeck, S., & Zila, P. L. (1997). Hands-on practice in learning to use software: A comparison of exercise, exploration, and combined formats. ACM Transactions on Computer-Human Interaction, 4(2), 169-196. Young, J. D. (1996). The effect of self-regulated learning strategies on performance in learner controlled computer-based instruction. Educational Technology Research and Development, 44, 17-27. Zimmerman, B. J. (1994). Dimensions of academic self-regulation: A conceptual framework for education. In D. H. Schunk, & B. J. Zimmerman (Eds.), Self-regulation of learning and performance (pp. 3-21). Hillsdale, NJ: Erlbaum.



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

EnDnoTEs 1



Unless otherwise noted, the Kruskal-Wallis (K-W) ANOVA by ranks and Friedman tests were used in this investigation, because they are conservative non-parametric tests that are best applied to ordinal and ratio data with small sample sizes (Maxwell & Delaney, 2000). The tests are based upon a chi-square (χ2 ) distribution.

2

3

4

5

Data for number of prior programming courses taken were missing for one student in Group 1. http://nasa1.ifsm.umbc.edu/learnJava/tutorLinks/TutorLinks.html http://nasa1.ifsm.umbc.edu/learnJava/savetext/TutorContent.pdf http://nasa1.ifsm.umbc.edu/IFSM413_613/

A Web-Based Tutor for Java™: Evidence of Meaningful Learning

aPPEnDiX a Following is the Java program that was taught by the tutoring system. The code is arbitrarily organized into 10 lines or rows. The method in Row 8, not needed with the default FlowLayout manager, was presented only to teach the application of a method on an object. Row 1: Row 2: Row 3: Row 4: Row 5: Row 6: Row 7: Row 8: Row 9: Row 10:

import java.applet.Applet; import java.awt.Label; public class MyProgram extends Applet { Label myLabel; public void init() { myLabel = new Label(“This is my first program.”); add(myLabel); myLabel.setVisible(true); } }



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

aPPEnDiX b rule-based Questions 1.

Which of the following lines most likely would be used to create a shorthand notation for the Frame class, which is built in to Java?

a. b. c. d.

import java.awt.frame; import java.awt.Frame.class; import java.awt.Frame; import java.awt.frame.class;

2.

Which of the following lines most likely would be used to construct an instance of the Button class?

a. b. c. d.

MyButton = new Button(“Hello”); myButton = new Button(“Hello”); myButton = button.class(“Hello”); MyButton = Button(“Hello”);

3.

Which of the following lines most likely would be used to add a Checkbox object to a container?

a. b. c. d.

Add(myCheckBox); Add(Checkbox); add(Checkbox); add(myCheckBox);

4.

Which of the following lines most likely overrides a method that is contained in the Applet class?

a. b. c. d.

public void stop(){ lines of Java code here } public void Stop{} { lines of Java code here } Public void Stop() ( lines of Java code here ) Public void stop() { lines of Java code here }

5.

Which of the following sequences is correct?

a. b. c. d.

declare a TextField object, construct a TextField object, add a TextField object to a container construct a TextField object, declare a TextField object, add a TextField object to a container declare a TextField object, add a TextField object to a container, construct a TextField object add a TextField object to a container, declare a TextField object, construct a TextField object



A Web-Based Tutor for Java™: Evidence of Meaningful Learning

6.

Given the line, public class MyTextArea extends TextArea {, which of the following statements is correct?

a. b. c. d.

TextArea is a subclass of MyTextArea. MyTextArea is a superclass of the extends class. TextArea is a superclass of MyTextArea. MyTextArea is a subclass of the Text class.

7.

Which one of the lines below declares myJFrame as a potential instance of the JFrame class?

a. b. c. d.

myJFrame extends JFrame. JFrame myJFrame: myJFrame JFrame; JFrame myJFrame;

8.

Given the following code: public class MyJFrame extends JFrame { … which one of the below would be the name of the Java file that contains this program?

a. b. c. d.

MyJFrame.Java MYJFrame.java MyJFrame.java MyJFrame.doc

9.

Which of the following lines most likely would add a JTextField object to a JPanel object?

a. b. c. d.

JPanel.add(JTextField); JPanel.add(myJTextField); myJPanel.add(JTextField); myJPanel2.add(myJTextField2);

10. A Java Applet program has two methods written in the class. The methods are not nested. What is the minimum number of braces, { and } added together, that are needed for this program? a. b. c. d.

9 6 3 4

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 2, edited by T. K. Shih, pp. 10-30, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).



0

Chapter XV

Personalisation in Web-Based Learning Environments Mohammad Issack Santally University of Mauritius, Mauritius Senteni Alain University of Mauritius, Mauritius

absTracT It is postulated that one of the main problems with e-learning environments is their lack of personalisation. This chapter presents a comprehensive review of the current work in the field and proposes a framework for research in promoting personalisation in Web-based learning environments. The concepts of adaptability, adaptivity and the limitations of completely adaptive systems are discussed. The conception of more interactive environments that are both adaptable and adaptive, which can assist the teacher in making interesting pedagogical decisions while tutoring in a virtual environment is proposed. Two versions of an algorithm that can be used to offer personalisation in the framework described are developed and discussed in this chapter. The algorithm is basically a method devised to select the most appropriate learning object from a pool of potential objects that exist in the repository.

inTroDUcTion Traditional distance education helped remove many barriers to education due to its relatively low price and high flexibility in the study modes. Students were given the opportunity to study at their own pace while working a full-time job. This

also motivated mature students to resume their studies without getting back to the school bench again. Nowadays, in this technology driven world, a new concept of distance education is emerging. Different interchangeable terms have been used to denote this concept: e-learning, online learning, Web-based learning and so forth. The concept of

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Personalisation in Web-Based Learning Environments

Web-based learning and the use of the Internet in teaching and learning have received increasing attention over the recent years. Using the Web as only a new kind of delivery medium for educational materials does not add significant value to the teaching and learning process. The integration of technology in learning needs to address the very important issue of enhancing the teaching and learning process, rather than just being seen as a new flexible delivery medium (Nichols, 2003). E-learning, if used effectively can help address the many shortcomings of the traditional distance education methods as well as the inherent problems the classroom teacher is faces on a daily basis with a classroom of learners with different learning and perceptual styles and competencies. The current research investigates the problems of personalisation in Web-based learning environments. It is in fact postulated that one of the main problems with e-learning environments is their lack of personalisation (Ayersman & Minden, 1995; Cristea, 2003; McLoughlin, 1999; Rumetshofer & Wöß, 2003). This research relates to Phase 3 of the University of Mauritius Learning Object Repository (UoM LOR) project (Santally, Govinda & Senteni, 2004) that investigates the possibility of adding adaptation elements in the authoring of courseware through the combination of learning objects. The elements that are focused upon relate to what is called the psychological traits of the students (Rumetshofer & Wöß, 2003) namely their learning styles, cognitive styles and controls. The Phase 3 of the UoM LOR project research has three main goals: 1.

2.

Devise an algorithm that will cater for personalised instruction by selecting the most appropriate learning object based on the existing student model. Develop the specifications of an adaptive instructional design framework to promote

3.

personalised and authentic activity-based learning experiences for the learner in a Web-based learning environment. Investigate the effects of adaptation to individual psychological traits (learning styles, cognitive styles and cognitive controls) of learners on the learning process through an implementation of the developed specifications using a learning object approach.

This chapter describes the adaptation framework and reports the observations made when two versions of the same adaptation algorithms are applied to a set of learning objects in a bid to identify the most appropriate one for any student with a particular individual profile. The student profile consists of a set of psychological attributes that are stored for the user namely the learning styles, cognitive styles and cognitive controls. The first version of the algorithm treats all modeled aspects to be of equal importance while the second one assigns importance weightage to each of the modeled factors. The aim is to observe the effects on the eventual selection of learning objects for the same students when the importance weightages are applied and randomly adjusted in the adaptation process.

a rEViEW oF lEarning sTYlEs, cogniTiVE sTYlEs, anD cogniTiVE conTrols The terms learning styles and cognitive styles have been often used interchangeably in literature. Jonassen and Grabowski (1993) distinguish between learning and cognitive styles by explaining that learning style instruments are typically self-report instruments, whereas cognitive style instruments require the learner to do a task which is then measured to some trait or preference.



Personalisation in Web-Based Learning Environments

learning styles

Types of learning styles

People prefer to learn in ways that are different from other people of the same class, culture or religion. This individual preference of how to learn is called the learning style preference. Education research and practice have demonstrated that learning can be enhanced when the instructional process accommodates the various learning styles of students (Buch & Bartley, 2002). Dunn (1996) postulates that gifted learners can learn proficiently without using their learning style preferences, however, low achievers significantly perform better when they capitalise on their preferences. Furthermore, a decade of research demonstrates that both low and average achievers earn higher scores on standardised achievement tests and attitude tests when taught through their learning style preferences (Dunn, Griggs, Olson, Gorman, & Beasley, 1995). Preferences for learning styles also change over time. However, during a period in which an individual has strong style preferences, that person will achieve most easily when taught with strategies and resources that complement those strategies (Dunn, 1996). It has been observed in learning style schools that many poor achievers do not function well under stress, but their stress appears sufficiently reduced after learning through their preferences to enable them to attain significantly higher scores on tests (Dunn et al., 1995). However, McLoughlin (1999) points out two different empirical findings showing that learning styles can hinder or enhance academic performance in several respects. In this respect, Dunn (1996) argues that teaching through learning styles is not enough and stresses the importance of the need for better assessment principles. Furthermore, other research on learning styles and achievement has shown that teaching students how to learn and how to monitor and manage their own learning styles is crucial to academic success (Atkinson, 1998; Biggs & Moore, 1993).

Dunn and Dunn Learning Styles



This model was first conceived in 1967 by Dr. Rita Dunn and further developed by both her and her husband, Dr. Kenneth Dunn. There are 5 main categories and 21 elements in considering a learning style when using their model. The Learning Styles Inventory (Dunn, 1996) would consist of the following: 1. 2. 3. 4. 5.

Environmental (Sound, Light, Temperature, Design) Emotional (Motivation, Persistence, Responsibility, Structure) Sociological (Self, Pair, Peers, Team, Adult, Varied) Physiological (Perceptual, Intake, Time, Mobility) Psychological (Global/Analytic, Hemisphericity, Impulsive/Reflective)

Due to the complexity of this learning-style model, each learner should be assessed to determine the best way to match teaching style to learner preference. This can be done by using the Productivity Environmental Preference Survey (PEPS) (Dunn, Dunn & Price, 1989). The PEPS is customised for use by adults and teachers. Students will receive high, medium or low scales in the five key categories of the Dunn and Dunn Learning-Style Model. These scores are neither good nor bad; they are unique to each individual. It is up to the learner to seek awareness from these scores, then create and modify his/her learning environment to best meet his/her needs. For example, if a student scored high on the sound element, studying with soft music playing in the background would increase his/her long-term memory retention and comprehension of the subject matter. For a learner who scored low on the sound element, this would have the opposite effect.

Personalisation in Web-Based Learning Environments

Grasha-Reichman Learning Styles This theory has to do with three personality dimensions: • • •

Competitive-collaborative Avoidant-participant Dependent-independent

The meanings of these dimensions are fairly self-evident based on the definitions of the words used in the terms. The definitions given below reveal how these students would prefer to learn in a Web-based environment.

Gregorc Learning Styles Gregorc and Butler’s theory organises learners on two dimensions: • •

Concrete/Abstract Sequential/Random

sions (abstract/concrete and active/reflective), which describe four abilities required to be an effective learner. Scores on the Kolb’s Learning Style Inventory identify the learner’s preferred style of receiving and organising information as an emphasis on abstractness over concreteness and on action over reflection (Ayersman & Minden, 1995). The Diverger is particularly adapted to viewing a situation from multiple perspectives and combines preferences for experiencing and reflecting. The Converger relies on common sense, is better suited to the practical application of ideas and combines preferences for thinking and doing. The Assimilator is viewed as a thinker who specialises in inductive reasoning and the formulation of theories. He/she combines preferences for thinking and doing. Finally, the Accommodator is more of a risk-taker who relies on intuitive trial and error approaches to problem solving and is highly adaptive to new situations (Ayersman & Minden, 1995).

The Honey and Mumford Approach In Gregorc and Butler’s work with perception and processing, two dichotomies are put into a matrix with the values concrete and abstract on the one axis, and the values sequential and random on the other (Dennis, 2002). Concrete refers to the ability to take in and adapt information based on facts and abstract refers to the ability to use imagination and intuition. Sequential covers the wish to obtain information in a linear process and random covers the preference for non-organised chunks of information. On this basis the learner can find his/her special learning potential characterised as abstract-sequential, abstract-random, concrete-sequential or concrete-random.

Kolb’s Learning Styles The Learning Style Inventory (LSI) by Kolb (1984) identifies four separate learning styles: Diverger, Converger, Assimilator, and Accomodator. These four styles of learning are assessed by two dimen-

A more promising alternative than the Kolb LSI may be a measure developed by Honey and Mumford (1986), named the learning styles questionnaire (LSQ). The Kolb model is the theoretical background to Honey and Mumford’s LSQ, which has four styles: Theorist, Activist, Reflector and Pragmatist (Figure 1). The Kolb model describes learning as a continuous process, which can be described as an endless loop. A learner can start anywhere on the cycle because each stage feeds into the next. A person could start, for example, at stage 2 by acquiring some information and pondering it before reaching some conclusions, stage 3, and deciding how to apply them, stage 4. People with Activist preferences, are well equipped for experience. People with the Reflector approach, with their predilection for mulling over data, are well equipped for reviewing and reflection. People with Theorist preferences, with their need to tidy up and have answers, are well



Personalisation in Web-Based Learning Environments

Figure 1. Honey and Mumford learning styles in Kolb cycle

equipped for concluding. Finally, people with Pragmatist preferences, with their liking for things practical, are well equipped for planning (Honey & Mumford, 1986).

cognitive styles

cognitive controls

Holists are global by initially creating broad interpretations of their environment while serialists are analytical by focusing on the details involved prior to making broad assumptions about their environment. According to Jonassen and Grabowski (1993), the serialist would prefer combining information linearly, and focus on small chunks of information at a time. Holists are described as the opposite of serialists, as being able to work on several aspects at the same time, having many goals and working on topics that span varying levels of structure.

The relevant literature reveals six main types of cognitive controls: field dependence/independence, cognitive flexibility, impulsivity and reflectivity, focal attention, category width and automisation (Ayersman & Minden, 1995). Cognitive controls influence and control an individual’s perception of environmental stimuli (Jonassen & Grabowski, 1993). Field dependent persons perceive objects as a whole, whereas field-independent persons focus on individual parts of an object. Cognitive flexibility determines a person’s ability to ignore distractions from his/her environment while focusing on relevant information at hand. Flexible individuals are able to effectively ignore irrelevant information while such materials would easily distract less flexible individuals (Jonassen & Grabowski, 1993). For more information on other cognitive controls, please see Ayersman and Minden (1995).



Information Organising (Serial/Holist and Analytical/Relational)

Information Gathering (Visual-Auditory-Kinaesthetic) This instrument has been developed by Barbe and Milone (1980) and describes the learner as visual, auditory and kinaesthetic. Persons with a visual preference tend to show a greater ability to analyse and integrate visual information, mentally convert nonvisual information into visual and show superior retention of mental images (Ayersman & Minden, 1995). An auditory learner, on the other

Personalisation in Web-Based Learning Environments

hand, would prefer to process information in the form of verbs, either written or spoken (Jonassen & Grabowski, 1993). Kinaesthetic learners prefer to process information through tactile means such as interactive media.

Personalisation in Web-based Environments The Concept of Aptitude Treatment Interaction Aptitude treatment interaction (ATI) research was developed as a way to find the best methods of instruction for the student population. Peck (1983) states that ATI is research correlating teaching methods with measures of student aptitudes finding that students may respond differently to a particular method depending on such variables as intelligence, learning style or personality. One major question that has developed from ATI research is the ability to transfer the findings by generalising the results to other groups. Four basic research designs standards, treatment revisions, aptitude growth and regression discontinuity have been recognised in ATI research. The standard design has been used in one form or another for educational purposes. The standard design is a simple randomised between-persons design where treatments are applied toward identified aptitudes to produce optimal outcomes. The outcomes must be measurable and must be measured prior to treatment and after treatment is applied. Because of ATI research, we now sometimes attempt to adapt instruction to the learner. The obvious implication of ATI research is, therefore, the adaptation of instruction to learners’ traits to maximise learning outcomes. Cronbach and Snow (1977) suggested the matching of instruction to traits at two levels: macro-adaptations, which match treatments to fit different classes of students, and micro-adaptations of treatments on a lesson-by-lesson, student-by-student basis. Macro-adaptation implies a multiple method ap-

proach to individualisation, the design of alternate treatments that engage different groups of students through different forms of information processing, whereas micro-adaptations focus on treatments to adapt the tasks and forms of instruction to meet more specific learner needs and abilities (Jonassen & Grabowski, 1993). Davis (1983) describes three approaches to the use of ATI’s to improve learning. The capitalisation approach urges the instructor to utilise with the student’s strengths. The compensation approach urges the instructor to provide a crutch if weakness is predicted, and then the remediation approach, in which the weakness is worked on until it is overcome.

The Concept of Adaptation (Adaptability vs. Adaptivity) The concept of “adaptation” is an important issue in research for learning systems. Systems that allow the user to change certain system parameters and adapt its behavior accordingly are called adaptable. Systems that adapt to the users automatically based on the system’s assumptions about the user needs are called adaptive. The whole spectrum of the concept of adaptation in computer systems is shown in Figure 2 (Patel & Kinshuk, 1997). Adaptivity in hypermedia systems to personalise the user’s experience with the system is not a new concept, and Brusilovsky (2001) describes three main types of adaptation that exist in Webbased hypermedia systems: content, navigation and layout. In adaptive hypermedia literature they are referred respectively as adaptive presentation and adaptive navigation support. Rumetshofer and Wöß (2003), on the other hand, postulate that in learning systems, adaptivity needs to cover more than what Brusilovsky (2001) proposes for Web-based hypermedia systems and propose what they call adaptation to psychological factors. These psychological factors are in fact the different factors such as cognitive styles, learning preferences and strategies that we already



Personalisation in Web-Based Learning Environments

Figure 2. Spectrum of the adaptation concept

mentioned. They propose the decomposition of the contents into a set of learning objects with an extension of the metadata to include the said psychological factors. Furthermore, they use an XML-XSLT based framework to provide a highlevel personalisation of courses to students. Cristea (2004) highlights the importance of connecting adaptive educational hypermedia with cognitive/learning styles on a higher level of authoring. She briefly reviews some systems and models that address the same issue but with different perspectives. The first system is TANGOW that actually implements the Felder-Silverman dimensions of learning styles. The system includes low-level authoring patterns such as learning material combination in AND, OR, ANY and XOR relations. The second system is AHA!, a low level tool with great flexibility based on IF-THEN rules adaptation model. The aim is to investigate how to incorporate high-level specifications deriving from learning styles especially those of fielddependent and field-independent styles into the low-level instances and structures as required by the AHA! system. Hong and Kinshuk (2004) develop a mechanism to fully model student’s learning styles and present the matching content, including contain, format, media type and so forth, to individual students, based on the Felder-Silverman Learning Style Theory. They use a precourse questionnaire to determine a student’s learning style or the stu-



dent may choose the default style and he/she is then provided with material according to his/her style. The efficiency of student learning with the prototype presented is however not yet tested. Magoulas, Papanikolaou, and Grigoriadou (2003) stress the importance of accommodating to individual differences when designing Webbased instructions. The authors propose a design rational and guidelines to implement adaptation strategies in such systems. Their model is based on the Kolb learning cycle and the Honey and Mumford (1986) learning styles. Wolf (2002) proposes iWeaver, an interactive Web-based adaptive learning environment. iWeaver uses the Dunn and Dunn learning style model and the Building Excellence Survey as an assessment tool to diagnose a student’s learning preferences. Instead of focusing on student’s learning preferences and to offer contents matching only a specific learning style of learners, iWeaver offers and encourages the trial of different media representations. It does not, however, adapt to the changing preferences of the learner. Ford (2000) explains that virtual environments enable a given information space to be traversed in different ways by different individuals using different roots and tools. He argues that cognitive styles are useful factors that can help in the personalisation of instruction in virtual environments. He furthermore argues the need of more robust student models to achieve better learning

Personalisation in Web-Based Learning Environments

systems design. He proposes virtual environments that could enable differential patterns and sequences of access to information that would suit the different types of students. Such access, according to Ford (2000), should be prescribed, autonomous or recommended.

ThE aDaPTaTion FramEWorK It has been stressed in the beginning that elearning environments need to add value to the teaching and learning process and constructivists who argue the importance of learner-centred approaches never take the risk of saying that there is no need for the teacher. Instead the teacher role is seen to change into a facilitator. A solution to such problems and that can be applied to real-world context to build hybrid adaptive learning environments where the human tutor is in constant interaction with the students. The system may have an adaptive engine supported by “enabling-agents” who help the teacher and student in better decision making. Inputs from the teacher will help update student models that can assist the agent in proposing better support to the actors in such environments. This idea is correlated to what Brusilovsky (2003) describes as meta-adaptation in adaptive environments. The need for human teachers in adaptive environments is also briefly argued by Kinshuk, Patel, Oppermann, and Russell (2001), as being the one who sets the context, selecting and scheduling other educational technologies, managing the curriculum and overseeing the learning progression. Brusilovsky and Nijhawan (2002) proposed the KnowledgeTree framework for adaptive e-learning based on distributed reusable learning activities. The KnowledgeTree framework is implemented in the form of a learning portal where the teacher works together with the adaptive system to select material to be presented to the learner. This is a step towards a hybridisation of

such learning environments. However, the role of the teacher is limited to selecting and sequencing materials to be presented and the system’s role is to propose additional instructional material to the student. Such a hybrid (or blended) model, therefore, should take into account the different actors in the learning environments, the tools each actor will have access to. For instance, a learner needs to access different tools to help him/her use his/her preferred learning strategies, the system has access to simple analysis tools to guide the student and to advise the teacher, the teacher needs to have access to pedagogical tools for tutoring; instructional models based on learning theories to devise appropriate instruction and learning activities and student models that will guide the system and teacher in adapting instruction to the students. The idea is to depart from classic intelligent tutoring, where the system (intelligent tutor) is limited in terms of knowledge, tutoring strategies and pedagogies, flexibility, lack of intuition and at the same time imposes restrictions on student actions and learning.

The student model The student model is an important part of the system, as it will contain the necessary individual adaptation attributes of the learner. From a recent survey at the University of Mauritius concerning students’ learning styles and cognitive styles (Santally, 2003), it was found that students can have preferences for one particular style, preference for more than one style and different levels of preferences for the different styles. The questionnaire consists of three sections (visual-auditory-kinesthetic) to determine the student’s strength in each field. The minimum that a student could score in any section is 12 and the maximum is 60. In the above cases, the students have more or less varied level of preferences. Furthermore, in the survey, the minimum score was 30 and maximum score 45. The same



Personalisation in Web-Based Learning Environments

Figure 3. Components of the student entity

observation was found in the Honey and Mumford (1986) LSQ that was used with the students to determine their learning preferences. It is clear that students would still learn if they were not given materials according to their preferences if the range of the scores was taken into account. Therefore, the goal of adaptation is in this context to present the student with the most suitable option in any given learning situation. The values, as shown in the Figure 3, would need to be rationalised since these will exist for the different elements that will be taken into account during the adaptation process. The process is simple: Effective (Belief) value for student preference = Actual Value/Σ (Auditory/Visual/Kinaesthetic) Figure 3 shows the student entity and the main attributes that will be stored in the student model. There are four main components: (1) cognitive style, (2) cognitive controls, (3) learning style and (4) performance. When a student registers in the system, his/her profile will also be stored so that this information may be used when he/ she starts studying in a particular course. The information to be stored in this profile can be obtained from cognitive style and learning style surveys and diagnostic operations by the teacher



or psychological experts. In an adaptive setting, the values of these attributes may need to change as the student gains more experience with the system and reaches a higher maturity level. The criteria on which the values will be adjusted are threefold: (1) the student’s perceptions history, that is the different feedback he/she gives when a learning object is presented to him/her, (2) his/her performance in the different milestone tests and (3) by the teacher who takes a decision based on the students interaction history with the system. Both the system and the teacher therefore provide adaptivity to the student’s learning.

The learning activity Psychologists and educators have long been interested in understanding how people learn, for the concept of learning is central to many different human endeavors (Shuell, 1986). Traditional conceptions of learning rely heavily on the notion of a durable change in behavior while from the cognitive perspective, the focus is principally laid on the acquisition of knowledge and knowledge structures rather than on behavior. Such approaches also postulate that learning is an active, constructive and goal-oriented process that is dependent upon the mental activities of the learner. Most cognitive conceptions of learning

Personalisation in Web-Based Learning Environments

Figure 4. Contextualising activity theory in an educational context

SUBJECT

RULES

OBJECT

COMMUNITY

OUTCOME

DIVISION OF LABOR

argue the need for “meaningful” learning to take place. Constructivists (and socioconstructivists) however, claim that meaningful learning occurs when the learner is placed in an authentic situation where he/she needs use his/her previous knowledge, percepts from the environments and cognitive processes to solve the problem at hand. Based on his/her experience, the learner constructs his/her own meaning through reflective activities. There is more and more claims that the real debate should not be centred around particular approaches to learning, but around how to achieve the real goal of education. That is, prepare students to apply the learned knowledge and skills in real world situations. The notion of “blended approach” (mixture of behaviorism, cognitivism, and constructivism) is also being widely emphasised to support such claims (Deubel, 2003; Schneider, 2003). Such a blended learning environment (Figure 4) can be seen from an activity theory perspective (Engeström, 1987) as consisting of a subject, a goal, a supporting community, rules governing the operation of the learning environment and a division of labor supporting the collective nature of the learning activities. The technology-supported learning environment can therefore be seen as a common instrumentality for a community

of learners (with the support of the teachers and resource persons) inquiring in and studying collectively a phenomenon or a social practice. Personalisation of this instrumentality would serve to develop the division of labor between learners who are inquiring in and learning about the same object of interest. The instruments in this case are the learning content, support materials for the activities, software tools and online forums.

The content model The learning objects approach to designing elearning courseware is not new. This is, however, a field still under research and the under-utilisation of existing learning object repositories is a major concern for many educators and researchers involved in this area (Santally et al., 2004). As with the learners’ preferences, a learning object may have components that suit visual learners better, but that also support kinesthetic learners up to a certain extent. Therefore, the same mechanism of adding belief values (in the range of 0-1) is applied to the content model to show the extent that a learning object can support a particular aspect. One constraint with the extension is that in the current setting there will be a need to compromise with the reusability of a particular learning object



Personalisation in Web-Based Learning Environments

Figure 5. Extension of learning object metadata (Adapted from Rumetshofer and Wöß 2003)

Figure 6. Learning objects in multiple representations for sequencing

since the belief value for the “pedagogical value” attribute will be restrictive to the context of utilisation of a particular learning object. A learning object may have a high pedagogical value when used in explaining the concept of Mechanics in an elementary physics course, but it might not have the same value if it were to be used in a Mechanics course for math. A section, for instance, will therefore be represented as a sequencing (in some cases, there may be loops, depending on the tutorial strategy) of learning objects. Each concept that will be illustrated in the section will consist of a series of learning objects with varying belief values for each component that has been added to the metadata description (Figure 5).

0

Concept 1 can be taught using LO1, LO2 or LO3. However, the selection of a learning object to teach a particular concept will depend on the student model and his/her current experience in the course. Sometimes a learning object can be presented more than once depending on the level of understanding of the student or on the tutoring requirements for this section. Different students will therefore have different pathways (Figure 6) to reach their learning goals and this brings the required flexibility and personalisation of the learning experience. A possible sequence for student X would be: [Concept 1, LO1]→[Concept 2, LO3]→[Concept 3, LO1]→[Concept n, LO2]

Personalisation in Web-Based Learning Environments

an algoriThm For PErsonalisED insTrUcTion

(0.3); Serial (0.2) Holist (0.8); Flexibility (0.7) Constriction (0.3)

The algorithm on which the system will decide which is the most appropriate learning object to be presented to the student will be based on kind of fuzzy approach. For each learning object that could be potentially presented to the student, the overall belief value or confidence factor (CF) will be computed and the one being closest to the student profile will be selected. The problem of adaptation in the current context is quite complex, since there are a number of parameters to consider for effective personalisation of the learning environment. Furthermore, the list of adaptation parameters is not exhaustive as other parameters can be progressively added. The algorithm reduces the complexity of the adaptation process and ensures that at any particular time the student is provided with the most appropriate learning object available. The algorithm has been designed in two variants.

LO3→ Visual (0.5) Auditory (0.2) Kinaesthetic (0.3); Field Dependent (0.1) Field Independent (0.9); Serial (0.2) Holist (0.8); flexibility (0.2) Constriction (0.8)

algorithm Variant 1 # 1: Consider student profile X (for simplicity reasons, only some attributes are chosen): X→ Visual (0.7) Auditory (0.2) Kinaesthetic (0.1); Field Dependent (0.6) Field Independent (0.4); Serial (0.8) Holist (0.2); Flexibility (0.4) Constriction (0.6) Now consider that there are three possible learning objects to be selected for presentation to the student: LO1→ Visual (0.4) Auditory (0.3) Kinaesthetic (0.3); Field Dependent (0.2) Field Independent (0.8); Serial (0.6) Holist (0.4); flexibility (0.6) Constriction(0.4) LO2→ Visual (0.2) Auditory (0.5) Kinaesthetic (0.3); Field Dependent (0.7) Field Independent

From the student profile, it is clear that student X has a visual preference, and is field dependent and is serialist and prefers constriction. Now, the task is to identify which of the three available learning objects would be most appropriate for the student. Step 1: Identify the LOs satisfying the most of the student’s strengths. •

LO1 (Visual [0.4] & Serial [0.6]), LO3 (Visual [0.5] & Constriction [0.8]), LO2 (Field Dependent [0.7])

Step 2: Compute overall belief of the values for each learning object by adding them together. •

LO1 [1.0], LO3 [1.3], LO2 [0.7]

Step 3: If overall belief values in more than one learning objects are equal, then the margin of error for each element is computed with respect to the student’s preferences. Otherwise, go to Step 5 directly. • •

LO1 [{0.4-0.7}+{0.6-0.8}] = -0.5 LO3 [{0.5-0.7}+{0.8-0.6}]= 0

Step 4: If margin of error is equal for the learning objects, then the system selects a learning object based on its previous history. That is, the evaluations and perceptions of other students in terms of its pedagogical relevance and usefulness, if any. If relevant statistics do not yet exist for the learning objects, a random selection is made.



Personalisation in Web-Based Learning Environments

Step 5: Select the learning object that has the maximum belief value. •

LO3

# 2: Now consider student profile Y (for simplicity reasons, only some attributes are chosen): Y→ Visual (0.4) Auditory (0.4) Kinaesthetic (0.2); Field Dependent (0.6) Field Independent (0.4); Serial (0.8) Holist (0.2); flexibility (0.5) Constriction (0.5) The difference with the profile of this student is that he/she is visual as well as auditory and he/she has no exact preference between flexibility and constriction. It could be either way for the student. The algorithm is slightly modified in this case since the student has preferences that are of the same strengths (magnitude): Now consider that there are three possible learning objects to be selected for presentation to the student: LO1→ Visual (0.5) Auditory (0.3) Kinaesthetic (0.2); Field Dependent (0.6) Field Independent (0.4); Serial (0.6) Holist (0.4); Flexibility (0.6) Constriction (0.4) LO2→ Visual (0.4) Auditory (0.4) Kinaesthetic (0.2); Field Dependent (0.3) Field Independent (0.7); Serial (0.4) Holist (0.6); Flexibility (0.6) Constriction (0.4) LO3→ Visual (0.5) Auditory (0.2) Kinaesthetic (0.3); Field Dependent (0.1) Field Independent (0.9); Serial (0.2) Holist (0.8); Flexibility (0.5) Constriction (0.5) Step 1: Identify the LOs satisfying the most of the student’s strengths. •



LO1 (Visual [0.5] & Serial [0.6] & flexibility [0.6]), LO2 (Visual [0.4] Auditory [0.4] &

flexibility [0.6]), LO3 (Visual [0.5] flexibility [0.5] & Constriction [0.5]) Step 2: Compute the margin of error for each element is computed with respect to the student’s preferences. • • •

LO1 [{0.5-0.4}+{0.3-0.4}+{0.6-0.6}+{0.60.8}+{0.6-0.5}+{0.4-0.5}] = -0.2 LO2 [{0.4-0.4}+{0.4-0.4}+{0.3-0.6}+{0.40.8}+{0.6-0.5}+{0.4-0.5}] = -0.7 LO3 [{0.5-0.4}+{0.2-0.4}+{0.1-0.6}+{0.20.8}+{0.5-0.5}+{0.5-0.5}] = -1.2

Step 3: If the margin of error of any two or more LOs from step 2 is equal, then compute the margin of error for only the values for which the LO(s) have maximum magnitude: • • •

LO1 [{0.5-0.4}+{0.6-0.6}+{0.6-0.5}] = 0.2 LO2 [{0.4-0.4}+{0.4-0.4}+{0.6-0.5}] = 0.1 LO3 [{0.5-0.4}+{0.5-0.5}+{0.5-0.5}] = 0.0

Step 4: If margin of error is equal again for the learning objects, then the system selects a learning object based on its previous history. That is, the evaluations and perceptions of other students in terms of its pedagogical relevance and usefulness, if any. This will be obtained through the pedagogical value attribute included in the learning object extended metadata. The lecturer/pedagogical designer will initially set the pedagogical value of a learning object and it will be adjusted based on perceptions and extent of use by students. If relevant statistics do not yet exist for the learning objects, a random selection is made. Go to step 5. Step 5: Select the learning object that has the maximum belief value •

LO1

Personalisation in Web-Based Learning Environments

algorithm Variant 2 The algorithm Variant 1 that has been devised assumes that all factors that are considered in the personalisation process are equally important and are to be taken into account. Some controls may affect a student learning experience more than another control. Therefore, if these differ in the learning objects metadata, there is a need to adapt the algorithm to cater for the factors that affect the student’s learning experience the most. The aim is to observe the effects on the eventual selection of learning objects for the same students when the importance weightages are applied and randomly adjusted in the adaptation process. In the weighted approach (Variant 2), it is assumed that one component of the student model may be dominant on the other and affects the student learning experience more than the other. This situation cannot be catered for in the normal approach since all factors are assumed to have equal importance in the personalisation process. In the weighted approach, each input (learning style strength) is multiplied by the corresponding adaptation weights and fed into the system. The summation function is used to compute the overall strength of the learning object according to rules already set by the algorithm (Variant 1 or Variant 2). An activation (transfer) function is then applied to the summation and the output is obtained (Figure 7). This is a function that takes a value and converts it to another value. The final value of the transfer function becomes the output

value. In this case the activation function used is a linear function (random): Output = 1.5 * (weighted_sum) -1 where sum is the magnitude of the learning object as computed in the normal approach. This approach is based on the neural network paradigm where the knowledge is stored in the weights, and not the inputs. In this case, after a pretest of the student’s preferences, adaptation takes place in terms of adjustment of the weights. It is worth mentioning, however, that there is no direct relationship of this approach to the neuron approach, at least at this stage of the process. The reason why a sigmoid transfer function is not used in the process is simple. The fact that the weights are in a range of 1-5 means that most of the time, the inputs will cancel each other and the effective value will not be significantly different. This is not applicable in the present research since the problem is not a neural network problem. The process is then repeated for each potential learning object and the one with highest output is retained and presented to the student. The instructional designer (in collaboration with the teacher/tutor), on an individual basis will assign weights for each student depending on his/her strengths. This will initially be based on the students’ responses to learning styles questionnaires and instructional designers’ experience. The weights will be adjusted depending on the students’ ratings of the learning objects, his/her

Figure 7. Computation of output for the weighted approach



Personalisation in Web-Based Learning Environments

feedbacks and progress on the learning process that is depending on the evolution of his/her personal profile. The weights basically establish a ratio for the different factors that are currently being modeled.

Xw→ Visual (0.7) Auditory (0.2) Kinaesthetic (0.1) Weight [2]; Field Dependent (0.6) Field Independent (0.4) Weight [3]; Serial (0.8) Holist (0.2) Weight [2]; Flexibility (0.4) Constriction (0.6) Weight [4]

Step 1: Identify the values of the learning for each factor modeled in the student.

Now applying the weighted version of the algorithm yields:

Step 2: Compute the margin of error for each element with respect to the student’s preferences for each learning object and multiply each value by the importance weightage (for each factor) and apply the activation function (for each factor) to obtain the effective value.

•

Step 3: If the margin of error of any two or more LOs from step 2 is equal, then compute the margin of error for only the values for which the LOs have maximum magnitude. Step 4-5: Same as variant 1 approach To observe how the second version of the algorithm works, the same student profile (X) and learning objects are used as in the normal approach. # 1: Consider student profile X (for simplicity reasons, only some attributes are chosen): X→ Visual (0.7) Auditory (0.2) Kinaesthetic (0.1); Field Dependent (0.6) Field Independent (0.4); Serial (0.8) Holist (0.2); Flexibility (0.4) Constriction (0.6) Now, assume that the student has been diagnosed by a pedagogical expert who assigns the following importance weights to the different factors that are modeled. The weights are assigned in a scale (1-5), (1) being the least important and (5) of utmost importance. The weighted student profile becomes:



•

LO1: o Sum = [{(0.4-0.7) * 2} +{(0.2-0.6)*3}+ {(0.6-0.8)*2}+ {(0.4-0.6)*4}] o Effective Value=1.5*sum-1= -5.5 LO2: o S u m = [{(0. 2 - 0.7 )*2}+{(0.70.6)*3}+{(0.2-0.8)*2)}+{(0.3-0.6)*4}] o Effective Value=1.5*sum-1= -5.65 LO3 o S u m = [{ ( 0 . 5 - 0 . 7 ) * 2 } + { ( 0 . 1 0.6)*3}+{(0.2-0.8)*2}+{0.8-0.6)*4}] o Effective Value=1.5*sum-1= -4.45

The learning object that will be selected is LO3. Although the same learning object was selected using the Variant 1 approach, there is not a necessary correlation between the two algorithms since everything depends on the weight assigned to each factor. In the coming sections, the weights will be randomly adjusted and the effect on the selection process will be observed. Consider the following example:

Initial Student Profile Visual (0.7) Auditory (0.2) Kinaesthetic (0.1) Weight [2]; Field Dependent (0.6) Field Independent (0.4) Weight [3]; Serial (0.8) Holist (0.2) Weight [2]; Flexibility (0.4) Constriction (0.6) Weight [4]

Personalisation in Web-Based Learning Environments

Table 1. Learning object selection with randomised weights Combination {[4][2][1][2]} {[1][3][4][1]} {[3][1][1][1]} {[2][1][4][4]} {[3][1][5][1]} {[2][5][1][2]} {[1][5][2][3]} {[4][1][5][2]}

Output LO3 LO1 LO3 LO3 LO1 LO2 LO2 LO1

Learning Objects LO1→ Visual (0.4) Auditory (0.3) Kinaesthetic (0.3); Field Dependent (0.2) Field Independent (0.8); Serial (0.6) Holist (0.4); flexibility (0.6) Constriction(0.4) LO2→ Visual (0.2) Auditory (0.5) Kinaesthetic (0.3); Field Dependent (0.7) Field Independent (0.3); Serial (0.2) Holist (0.8); Flexibility (0.7) Constriction (0.3) LO3→ Visual (0.5) Auditory (0.2) Kinaesthetic (0.3); Field Dependent (0.1) Field Independent (0.9); Serial (0.2) Holist (0.8); flexibility (0.2) Constriction (0.8) Randomising the weight combinations to be applied and corresponding output computed yields different learning objects for different weight combinations (Table 1).

DiscUssion It is clear that contrary to the outcomes from the first version of the algorithm (Appendix 1) and from Table 1, there are variations in the learning object selection process when the weight combinations are altered for any particular student profile. For the same student profile, using 16 different combinations showed that each learning object had a chance of being selected depending on the value

Combination {[3][5][1][4]} {[1][1][1][5]} {[1][1][1][1]} {[4][4][4][4]} {[1][2][3][5]} {[4][5][3][4]} {[5][3][2][1]} {[3][4][1][5]}

Output LO2 LO3 LO1/LO3 LO1/LO3 LO3 LO3 LO3 LO3

of the weights. One thing that is worth pointing out is the neutralising effect of combinations [1,1,1,1] to [5,5,5,5]. In fact, there is no difference in selection whenever any of these combinations is used. The ambiguity lies in the fact that a combination of [1,1,1,1] means the student has weak preferences linked to learning/cognitive styles. This suggests that providing adaptation to learning styles of such learners may not be the right idea because some other factors may be more predominant in affecting their learning process. On the other hand, while the learning object selected will still be the same, a learner with profile weights [5,5,5,5] means he/she is a versatile learner who has individual preferences and each factor is equally important in ensuring a smooth learning process for the said learner. This factor also reveals a weakness in the first variant of the algorithm since the weighted version becomes equivalent to the normal version in such combinations. This means that the first version of the algorithm had two inherent assumptions. The first assumption is that it is taken for granted that learning/cognitive styles of the students already plays a predominant role in the learning process of each student, and the second one assumes that all factors modeled are equally important for each student. It would be indeed interesting to treat such cases independently as the outcomes may be completely different for these students and this may cause bias in related and subsequent studies. Another implication related to the second version concerns the adjustment of values as the stu-



Personalisation in Web-Based Learning Environments

dent interacts with the system to provide efficient adaptation. The task here is to decide whether to change the weights, or the individual ratings or both sets of values. There are two levels of updates that can be done to provide adaptation. For instance, we can update the visual preference value of the student if we think his/her visual preference is increasing over his/her auditory preference or just adjust the cognitive style weights if we think the student is developing this component further than, say, the cognitive controls component. The task of monitoring a student progress and the decision making process for changing the attributes of the student is a time-consuming one for the teacher. A pedagogue may be needed to assist the teacher in the process. At the same time, this definitely puts constraints over the number of students a teacher can coach at a time. To help the teacher in the process, a system agent will provide him/her with statistically interpreted/grouped data about the student actions and interaction history with the system. Assuming the teacher’s decisions to be correct, an automated model of updating the students’ attributes can be conceived in the long run. However, this approach will seem to contradict the very basis of the research presented in this chapter, that is, about emphasis on the teacher’s role in the adaptation process. A solution to this problem is then purely pedagogical, decreasing the class size responsibility for tutoring. From experience, the instructional design process for e-learning courseware has always been a fuzzy issue from writing the courseware, designing activities and developing of user-centred interactive Web sites. Adding a method for adaptation in the process does no less than adding complexity to it and a teacher would never be able to handle that process alone. A team consisting of the teacher, tutor, instructional designer, a system agent and the student need to work collectively to make the adaptation process successful. The task of rating learning objects and subsequently updating these values and the attributes of the



student is a very difficult task and clear rules and guidelines need to be established for these. This suggests a need to apply a relevant instructional design methodology that will take into account students’ preferences and individual characteristics in the courseware authoring process. Of course, the idea here is not to reinvent the wheels by proposing a new instructional design method, but existing ones can be adapted to suit the approach discussed here. While a learning object approach to instructional adaptation is favored in this context, there is also the belief here that learning objects are by themselves not teaching objects and that there is no guarantee that learning will take place in contextually meaningful ways. Most adaptation, learning styles and related computer/Web-based learning research is based on the fact that either the system is a kind of an intelligent tutor, or by presenting same material through different ways and finally exposing the student to a test, efficient adaptation has taken place or learning has taken place. The problem behind this perspective is that learning through assimilation of facts has taken place, but there is no guarantee that the student has developed and acquired competencies that will be applicable in real contexts. The way learning is defined will therefore definitely affect the adaptation method. The method presented here is, therefore, the first part of a twofold process. It is assumed that students will be engaged in a collective activity, say, construction of a vapor-propeled board to illustrate the energy conversion phenomena. To achieve their individual parts, each student will be exposed to learning materials as per his/her individual preferences. When the parts of each student are assembled together, the project is realised. In this way, the students have acquired competencies by using their preferred learning strategies/styles and then share this experience with peers of the group and create the project in a collective manner. This approach of individualised collective work can be beneficial to both the individual and the group.

Personalisation in Web-Based Learning Environments

conclUsion This chapter highlights the need to extend current adaptation frameworks by including the human teacher in diagnosing student abilities, cognitive styles and preferred learning strategies to offer improved personalisation in Web-based learning environments. The problem of personalisation is therefore not only restricted to the student, but also applies to a whole activity system in innovative educational contexts where a wide range of pedagogies are supported. The idea is to demark from classic intelligent tutoring where the system (intelligent tutor) is limited in terms of knowledge, tutoring strategies and pedagogies, flexibility, lack of intuition and at the same time imposes restrictions on student actions and learning.

rEFErEncEs Atkinson, S. (1998). Cognitive style in the context of design and technology work. Educational Psychology, 18(2), 183-194 Ayersman, D. J., & Minden, A. (1995). Individual differences, computers and instruction. Computers in Human Behaviour, 11(3-4), 371-390. Barbe, W., & Milone, M. (1980). Modality. Instructor, 89(6), 44-46. Biggs, J. B., & Moore, P. J. (1993). The process of learning (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. Bruvilovsky, P. (2001). Adaptive hypermedia. User Modeling and User-Adapted Interaction, 11, 87-110. Bruvilovsky, P. (2003). Adaptive navigation support in educational hypermedia: The role of student knowledge level and the case for metaadaptation. British Journal of Educational Technology, 34(4), 487-497.

Brusilovsky, P., & Nijhawan, H. (2002, October). A framework for adaptive e-learning based on distributed re-usable learning activities. In M. Driscoll & T. C. Reeves (Eds.), Proceedings of World Conference on E-Learning, E-Learn 2002 (pp. 154-161). Montreal, Canada. Buch, K., & Bartley, S. (2002). Learning style and training delivery mode preference. Journal of Workplace Learning, 14(1), 5-10. Cristea, A. (2003). Adaptive patterns in authoring of educational adaptive hypermedia. Educational Technology & Society, 6(4). Retrieved July 18, 2006, from http://ifets.ieee.org/periodical/6_4/1. pdf Cristea, A. (2004, February). Adaptive and adaptable educational hypermedia: Where are we now and where are we going? In Proceedings of Webbased Education, Innsbruck, Austria. Retrieved November 8, 2005 from http://wwwis.win.tue. nl/~acristea/HTML/ftp_paper_chrono.html Cronbach, L., & Snow, R. (1977). Aptitudes and instructional methods: A handbook for research on interactions. New York: Irvington. Davis, G. A. (1983). Educational psychology. New York: Allyn & Bacon. Dennis, W. (2002) Applying what we know: Student learning styles. Retrieved July 18, 2006, from http://www.csrnet.org/csrnet/articles/student-learning-styles.htm Deubel, P. (2003). An investigation of behaviorist and cognitive approaches to instructional multimedia design. Journal of Educational Multimedia and Hypermedia, 12(1), 63-90. Dunn, R. (1996). How to implement and supervise a learning style program. VA: ACSD. Dunn, R., Dunn K., & Price, G. (1989). Productivity environmental preference survey. Lawrence, KS: Price Systems.



Personalisation in Web-Based Learning Environments

Dunn, R., Griggs, S. A., Olson, J., Gorman, B., & Beasley, M. (1995). A meta-analytic validation of the Dunn and Dunn learning styles model. Journal of Educational Research, 88(6), 353-361. Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental research. Helsinki: Orienta-Konsultit. Ford, N. (2000). Cognitive styles and virtual environments. Journal of the American Society for Information Science, 51(6), 543-557. Honey, P., & Mumford, A. (1986). Using your learning styles. Maidenhead, Berkshire: Honey Publications. Hong, H., & Kinshuk. (2004, June). Adaptation to student learning styles in web-based educational systems. In L. Cantoni & C. Mcloughlin (Eds.), Proceedings of ED-MEDIA 2004 - World Conference on Educational Multimedia, Hypermedia & Telecommunications (pp. 491-496), Lugano, Switzerland. Jonassen, D. H., & Grabowski, B. (1993). Individual differences and instruction. Computers in Human Behaviour, 11(3-4), 371-390. Kinshuk, Patel, A., Oppermann, R., & Russell, D. (2001). Role of human teacher in Web-based intelligent tutoring systems. Journal of Distance Learning, 6 (1), 26-35. Kolb, D. (1984). Experiential learning. Englewood Cliffs, NJ: Prentice Hall. Magoulas, G., Papanikolaou, K., & Grigoriadou, M. (2003). Adaptive Web-based learning: Accommodating individual differences through system’s adaptation. British Journal of Educational Technology, 34(4). Retrieved July 19, 2006, from http://www.dcs.bbk.ac.uk/~gmagoulas/bjet.pdf Mcloughlin, C. (1999). The implications of research literature on learning styles for the design of instructional material. Australian Journal of Educational Technology, 15(3), 222-241. Retrieved



July 19, 2006, from http://www.ascilite.org.au/ ajet/ajet15/mcloughlin.htm Nichols, M. (2003). A theory for e-learning. Educational Technology & Society, 6(2), 1-10. Patel, A., & Kinshuk. (1997). Intelligent tutoring tools in a computer integrated learning environment for introductory numeric disciplines. Innovations in Education and Training International Journal, 34(3), 200-207. Peck, M.L (1983). Aptitude treatment interaction research has educational value. In Proceedings of selected research paper presentations at the 1983 convention of the Association for educational communicators and technology and sponsored by the research and theory division. (pp. 564-622). New Orleans: Association for Educational Communications & Technology. Rumetshofer, H., & Wöß, W. (2003). XML-based adaptation framework for psychological-driven e-learning systems. Educational Technology & Society, 6(4). Retrieved July 19, 2006, from http://ifets.ieee.org/periodical/6_3/4.pdf Santally, M. (2003). Individual instruction and distance learning: Application of learning and cognitive styles. Malaysian Journal Of Distance Education, 5(2), 15-26. Santally, M., Govinda, M., & Senteni, A. (2004). Reusable learning object aggregation for e-learning courseware development at the University of Mauritius. International Journal of Instructional Technology and Distance Learning. Retrieved July 19, 2006, from http://www.itdl.org/Journal/ Jul_04/article02.htm Schneider, D. (2003). Conception and implementation of rich pedagogical scenarios through collaborative portal sites: Clear focus and fuzzy edges. In Proceedings of the International Conference on Open & Online Learning, ICOOL 2003, Mauritius. Retreived December 5, 2004 from http://icool. uom.ac.mu/2003/papers/schneider.php

Personalisation in Web-Based Learning Environments

Shuell, T. J. (1986). Cognitive conceptions of learning. Review of Educational Research, 56(4), 411-436.

Wolf, C. (2002). iWeaver: Towards an interactive Web-based adaptive learning environment to address individual learning styles. European Journal of Distance Learning. Retrieved July 19, 2006, from http://www.eurodl.org/materials/contrib/2002/2HTML/iWeaver



0

SP 1 LO1 LO2 LO3 SP 2 LO1 LO2 LO3 SP 3 LO1 LO2 LO3 SP 4 LO1 LO2 LO3 SP 5 LO1 LO2 LO3 SP 6 LO1 LO2 LO3 SP 7 LO1 LO2 LO3 SP 8 LO1 LO2 LO3

Serialist 0.1 0.28 0.23 0.3 0.97 0.92 0.65 0.52 0.26 0.67 0.38 0.81 0.69 0.43 0.18 0.04 0.47 0.64 0.1 0.01 0.68 0.01 0.81 0.63 0.93 0.95 0.63 0.49 0.43 0.38 0.5 0.85

COGNITIVE STYLES Visual Auditory Kinesthetic 0.15 0.6 0.25 0.2 0.64 0.16 0.13 0.32 0.55 0.3 0.24 0.46 0.58 0.28 0.14 0.35 0.51 0.13 0.51 0.37 0.12 0.28 0.42 0.3 0.58 0.28 0.14 0.57 0.32 0.11 0.21 0.67 0.11 0.18 0.67 0.14 0.33 0.18 0.49 0.71 0.14 0.15 0.36 0.22 0.42 0.18 0.6 0.21 0.24 0.56 0.2 0.23 0.38 0.39 0.13 0.23 0.64 0.56 0.18 0.26 0.4 0.23 0.37 0.23 0.39 0.38 0.29 0.23 0.48 0.39 0.26 0.35 0.26 0.35 0.39 0.28 0.17 0.55 0.16 0.71 0.13 0.64 0.26 0.11 0.61 0.17 0.22 0.71 0.15 0.13 0.41 0.35 0.24 0.17 0.35 0.48

Holist 0.9 0.72 0.77 0.7 0.03 0.08 0.35 0.48 0.74 0.33 0.62 0.19 0.31 0.57 0.82 0.96 0.53 0.36 0.9 0.99 0.32 0.99 0.19 0.37 0.07 0.05 0.37 0.51 0.57 0.62 0.5 0.15

FD 0.18 0.55 0.79 0.11 0.52 0.69 0.2 0.5 0.25 0.36 0.71 0.96 0.61 0.9 0.49 0.56 0.73 0.18 0.09 0.83 0.3 0.12 0.22 0.81 0.59 0.22 0.05 0.57 0.18 0.23 0.44 0.68

FI 0.82 0.45 0.21 0.89 0.48 0.31 0.8 0.5 0.75 0.64 0.29 0.04 0.39 0.1 0.51 0.44 0.27 0.82 0.91 0.17 0.7 0.88 0.78 0.19 0.41 0.78 0.95 0.43 0.82 0.77 0.56 0.32

Flexibility Constriction 0.1 0.9 0.66 0.34 0.34 0.66 0.23 0.77 0.45 0.55 0.94 0.06 0.43 0.57 0.85 0.15 0.13 0.87 0.31 0.69 0.95 0.05 0.22 0.78 0.2 0.8 0.1 0.9 0.98 0.02 0.56 0.44 0.06 0.94 0.24 0.76 0.84 0.16 0.8 0.2 0.9 0.1 0.29 0.71 0.17 0.83 0.24 0.76 0.62 0.38 0.85 0.15 0.74 0.26 0.96 0.04 0.1 0.9 0.45 0.55 0.42 0.58 0.52 0.48

COGNITIVE CONTROLS Expert Belief 3 5 3 5 3 1 5 3 5 4 3 5 2 4 3 3 2 1 4 3 2 1 5 5

(1.61) --- (-0.60) (1.73) --- (-0.69) (1.02) --- (-1.17) (1.90) --- (-0.71) (0.62) --- (-1.77) (0.78) --- (-1.75) (1.80) --- (-0.21) --- (-1.48) (0.56) --- (-0.62) (0.76) --- (-1.08) (0.90) --- (-1.38) (1.82) --- (-0.56) (0.88) --- (-1.27) (1.59) --- (-0.63) (1.02) --- (-1.23) (2.35) --- (0.04) (1.37) --- (-0.98) (1.53) --- (-0.40) (2.65) --- (-0.25) (2.05) --- (-0.85) --- (-1.78)

Ped. Value

(1.36) --- (-1.07) (1.43) --- (-1.26) (2.36) --- (-0.62)

Computed A1 --- A2

Personalisation in Web-Based Learning Environments

aPPEnDiX a. DaTa sETs anD rEsUlTs WhEn TEsTing ThE VarianT 1 oF algoriThm

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 4, edited by S.K. Chang and T. K. Shih, pp. 15-35, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).



Chapter XVI

Implementation and Performance Evaluation of WWW Conference System for Supporting Remote Mental Health Care Education Kaoru Sugita Fukuoka Institute of Technology (FIT), Japan Giuseppe De Marco Fukuoka Institute of Technology (FIT), Japan Leonard Barolli Fukuoka Institute of Technology (FIT), Japan Noriki Uchida Global Software Corporation, Japan Akihiro Miyakawa Nanao City, Ishikawa Prefecture, Japan

absTracT Information technology (IT) has changed our lives and many applications are based on IT. IT can be helpful for remote mental health care education. Because there are very few mental health care specialists, it is very important to decrease their moving time. But it is not easy to use the conventional TV conference systems for ordinary people, mental health care specialists, and their students because they

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Implementation and Performance Evaluation of WWW Conference System

are not computer specialists. For this reason, we have developed a WWW conference system. Our system can communicate between the mental health care specialists and their students by using the live video on WWW browser. In this paper, we show the implementation and the evaluation of proposed system. The experimental results over the Internet show that our system can be used for real time communication between Fukuoka, Ishikawa, and Iwate prefectures.

inTroDUcTion Recently in Japan, mental health care has become a very important issue because there are many people suffering from mental problems. Also, there are only a few specialists and researchers to deal with these problems. For example, bullying at school is one of the big problems worrying teachers in Japan today. Also, there are other problems such as refusal to go to school and school violence. However, there are few counselors in Japanese schools. It should be noted that many of these specialists are also doing other works. In general, one mental health care specialist should take care for many patients and they want to see their face. They should take care of not only the counseling but also the individual aftercare. Information technology (IT) has changed our lives. People can communicate between themselves and students can study various courses at anywhere and anytime using the Internet. IT can be helpful for mental health care, education, aftercare, and counseling for patients and their families. Because there are very few mental health care specialists, it is very important to decrease their moving time. Also, it is very important to see the facial expression and to talk to people for mental health care education, aftercare, and counseling. For this reason, the video and voice are needed. Recently, because of the use of ADSL and FTTH (NTT West, 2004) many people can use several Mbps on the Internet. Therefore, many people can use the streaming live video and TV conference. For example, the streaming live video is used for the tourist attractions broadcasting



(Fuji-shi, 2005; Kumamoto-shi, 2001; Sapporoshi, 2004), assembly broadcasting (The House of Councilors, 2005; The House of Representatives, 1999;), and public information broadcasting (The ministry of public management, 2004). In these live streaming video methods, the image refreshed at regular intervals over WWW (Fuji-shi, 2005; Kumamoto-shi, 2001; Sapporo-shi, 2004) and RealPlayer or Windows Media Player (The House of Councilors, 2005; The House of Representatives, 1999; The Ministry of Public Management, 2004) are used. Some TV conference systems are used as special hardware (NEC Engineering, 1996; Polycom, 2004; SONY, 2004; VTEL, 2004) and in other systems are used as application software or PC connected to the existing cameras (Advanced Solutions, 2001, 2002; Hitachi Hybrid Network, 2004; Microsoft, 2004; Visual Nexus, 2004). When using the special hardware, the users need to connect to the Internet and set up the hardware in each spot before using the system. But, it is difficult to use the special hardware in counseling because the users are usual people. While, when using the application software and PC is connected to the existing camera, the users need to connect the PC and camera, setup the PC, and install the application software in each spot before using the system. Also, it is difficult to introduce these systems for mental health care because the mental health care specialists and their student are not computer specialists. In order to realize a remote mental health care education, we have developed a WWW conference system. Our system is able to support the communication between the mental health care

Implementation and Performance Evaluation of WWW Conference System

specialists and their students. Also, our system can provide the communication between the mental health care specialists, patients, and their families by using the live video on WWW browser, point to point communication, point to multi-point communication, and multi-point to multi-point communication. The organization of this chapter is as follows. In the next section, we will introduce the related works. Next, we describe the WWW conference system and explain system architecture. The conference types are treated in the following section. After that, we show the connection management for each conference type and the flow of WWW conference system. Then, we show implementation of our proposed system by using Macromedia Flash and Macromedia Flash Communication Server and provide the experimental results. Finally, we give some conclusions and future works.

rElaTED WorK Existing communication systems are usually realized on the LAN environment and leased lines. Many remote education systems are proposed in Inoue, Okada, and Matsushita (1997), Mori, Oyabu, Nomura, and Oshita (1992), Sakiyama, Ohono, Mukunoki, and Ikeda (2001), and Wakahara (1998). In Inoue et al. (1997) and Wakahara (1998), the multi-point to multi-point communication is used. While in Sakiyama et al. (2001), a video stream selection according to lecture context is proposed. In these systems, the remote users can communicate between them using the live video. Recently, the live video communication systems have become very popular and the network speed is increased very fast. Therefore, the video conference systems can be applied in the medical field. There are many video conference systems such as the medical tele-consultation support system using super high definition imaging sys-

tem (Yamaguchi et al., 2001), remote medical support system using the transcoding function (Kawamura, Maita, Hashimoto, & Shibata, 2003), and care communications service between hospitalized patients and their families (Abiko, Iijima, Koyama, Kamibayashi, & Narita, 2003). In these studies, the system was implemented only for intranet or leased lines. Some other communication systems using live video for remote learning systems (Maeda et al., 1997; Miyoshi, Okanaga, Kou, & Kondo, 2000) and streaming video (Kato, Jiang, & Hakozaki, 2003) has been proposed. In the multimedia communication environment for distance learning proposed in Maeda et al. (1997) are given the experimental results for remote learning using intranet and Internet environments. In this study is shown that an efficient remote learning can be achieved by using 600 Kbps bandwidth. Now, Internet users can have more bandwidth by using ADSL and FTTH. And the live video over the Internet is possible. In Kato et al. (2003), the authors propose a streaming video system for best-effort networks using adaptive QoS control rules to improve the satisfaction of streaming video services. However, these systems can’t be used by usual people and they need special hardware and software. These systems use point to point communication, so they can not be used for multipoint to multi-point communication. In order to overcome these problems, we have developed a WWW conference system, which uses multi-point to multi-point communication and can transmit live video over the Internet.

ProPosED WWW conFErEncE sYsTEm The proposed WWW conference system is shown in Figure 1. If the user has a video camera and PC connected to the Internet, the user can communicate to the remote users by using live video on



Implementation and Performance Evaluation of WWW Conference System

Figure 1. WWW conference

Ishikawa Nanao-shi Fukuoka Fukuoka Institute of T echnology

Iwate Netbridge Company

Internet

WWW. Using our system, the user does not need to install and setup the application software. He just only needs to access the WWW conference server by using the WWW browser. The concept of WWW conference is shown in Figure 2. The user can use various cameras such as the WWW camera, digital video camera and robot camera. But, the camera should be identified and connected to the PC. Our system is composed of the WWW conference client and the WWW conference server. When the user accesses the WWW conference server by using a WWW browser, the WWW conference client is downloaded to the PC. After downloading the WWW conference client, the system provides the following functions: 1. 2. 3. 4.



User authentication Live video function Video record function Quality setting function

The user authentication function is provided by user ID and password. When the user enters into the WWW conference server, the user inputs its user ID and password. The user ID and password are compared with the registered ones in the WWW conference server. When the user ID and password do not match, the user login fails and the user needs to input again its user ID and password. It should be noted that the user login may fail when the WWW conference server has already logged on the maximum number of users. The live video function provides the video data and voice data. The video data and voice data are sent and received by the video stream. They are encoded by H.263. Our system provides three types of conference using this function. In order to realize the live video, the WWW conference client has the connection management function, the video encode function, and the video decode function. The WWW conference server has the connection management function.

Implementation and Performance Evaluation of WWW Conference System

Figure 2. WWW conference system WWW Conference Client -Send and receive the live video -Play the storage video -A djust the video quality Macintosh

WWW Conference Client -Send and receive the live video -Play the storage video -A djust the video quality

Digital video camera

R obot camera

PC

WWW Conference Client -Send and receive the live video -Play the storage video -A djust the video quality PC

Web Camera

R S232C IE E E 1394

USB

NT SC

Internet -Provide the WWW conference -User authentication -Connection management -R ecord the live video WWW Conference Server

The video record function records the live video stream and plays the recorded video stream on the WWW conference server. The user can record the live video stream and play the recorded video stream by using the WWW conference client. The recorded video stream is stored in the WWW conference server. The recorded video stream is used for learning the existing cases of counseling and replaying the instructions for mental health care. The quality setting function modifies the video quality for local WWW conference client and remote WWW conference client. The local WWW conference client can modify the video size and frame rate. But, the remote WWW conference client can modify only the frame rate. When the PC does not have enough CPU power or network bandwidth, the voice data are delayed and video data experiences frame loss and frame delay. However, our system can guarantee the quality of service such as “Frame rate is high and video size is small” and “Frame rate is low and video size is large” when the user modifies the video quality using this function.

sYsTEm archiTEcTUrE In order to implement our system, we propose a system architecture, which is organized by the WWW conference client and the WWW conference server as shown in Figure 3.

WWW conference client Presently, various network environments exist, such as broadband network and narrowband network. The bandwidth varies from 10 Kbps to 100 Mbps. In order to realize the communication between various network environments, the WWW conference client is organized by user interface (UI), authentication manager (AM), User Manager (UM), WWW conference manager (WCM), connection manager (CM), shared object manager (SOM), video I/O manager (VI/OM), video quality manager (VQM), and video stream manager (VSM). The UI provides the following functions: 1.

Insertion of user ID and password



Implementation and Performance Evaluation of WWW Conference System

Figure 3. System architecture WWW Conference Client

WWW Conference Server

User Interface

Authentication Manager

WWW Conference Manager

Video I/O Manager

User Manager

Video Quality Manager

Shared Object Manager

Connection Manager

Video Stream Manager

Video Recording Manager

Video Stream Manager

Network Interface

WWW Conference Manager User Manager

Connection Manager

Authentication Manager

Shared Object Manager

Network Interface

Internet Message

2. 3. 4.

Object

Video stream

Setup of video input and output device Display of live video Modification of video quality

The AM authenticates the user ID and password by using the registered user ID and password on the WWW conference server. When the AM fails for the authentication, the AM requests the user ID and password to the user again. When the AM authentication is a success for the authentication, the AM sends the user ID to the UM. The UM manages the system users, the video quality for each video stream and the name of the recorded video stream. When maximum number of users has been logged in to the WWW conference server, the UM rejects other users. The WCM sets up the number of maximum users and provides the user interface. The maximum number of users and the user interface are decided by the conference type. The CM manages the connections of the other WWW conference clients participating in the WWW conference. The CM converts

256

the number of maximum users to the number of maximum connections. The SOM gets the shared object from the WWW conference server. The recorded video stream is played by using the video name. We use the shared object to record the name of video stream on the WWW conference server. When the user wants to record the live video stream through the UI, the SOM adds the name of video stream as the shared object to the WWW conference server. The VI/OM inputs the video data and sound data from the camera and the video stream is selected by the WCM. The VI/OM displays the video data and sound data from the video stream and the camera. Also, the VI/OM sends the video data and sound data to the VQM. The VQM decides the frame rate and the size of video by using the identified video quality. The VSM sends and receives the video data and voice data as the video stream to the WWW conference server. The other WWW conference clients operate in the same way.

Implementation and Performance Evaluation of WWW Conference System

WWW conference server

TYPE oF conFErEncEs

In order to realize the WWW conference and record the live video stream, it is needed to manage the video stream including live video stream and recorded video stream. So, the WWW conference server is organized by the AM, WCM, UM, CM, SOM, Video Record Manager (VRM), and VSM. The AM manages the user ID and password. It compares the user ID and password with the registered user ID and password when the AM requests the authentication from the WWW conference client. The UM manages the users inside the system, the video quality for each video stream and the name of the recorded video stream. The WCM manages the user and video stream based on type of conference. The CM manages the connection of the WWW conference clients. Then, it decides the number of maximum connections by using the number of maximum users. When the WWW conference server has already the maximum number of users, the CM rejects the new connection. The SOM stores the shared object, which is the name of video stream on the WWW conference server. The shared object remains even after disconnecting the WWW conference client. The VRM records and manages the video stream on the WWW conference server. The VSM manages the video stream between the WWW conference clients and the WWW conference server.

For the remote mental health care education are needed various conference types. We introduce three conference types because our system is used for the counseling, aftercare, and education.

Figure 4. Point to point conference

Figure 5. Multi-user conference

Point to Point conference The point to point conference is used for counseling the patients. The point to point conference provides functions for point to point communication as shown in Figure 4. This type displays the local live video stream and the remote live video stream. The user can record the local live video stream and play the recorded video stream. Using the point to point conference, the patient can take the counseling at home.

multi-User conference The multi-user conference is used for aftercare. The multi-user conference provides functions for the multi-point to multi-point communication as shown in Figure 5. In the multi-users conference, the users can communicate in a distributed way at the same time. This is why the local video stream and the multiple remote live video streams can be display at the same time. By using the multi-user conference, the patient and their family members can talk together with psychology specialists even if they are not at the same place.



Implementation and Performance Evaluation of WWW Conference System

Figure 6. Broadcast conference

broadcast conference The broadcast conference is used for health care education. It provides broadcast functions to broadcast the live video to many users as shown in Figure 6. This method is used for live broadcast (The House of Councilors, 2005; The House of Representatives, 1999). The broadcast conference uses two types of WWW conference clients: the sender and receiver clients. The sender client displays the local live video, selects the sending channel, sends the live video stream to receiver client, and records the local live video stream. The receiver client selects the receiver channel, receives and displays the video stream and plays the recorded video stream. Using the broadcast conference, the students can learn the clinical psychology, the family and patient can learn the mental health care at home.

connEcTion managEmEnT In order to provide the live video for distributed users, the video data and voice data are transmitted in the video streams. The number of video streams is decided based on the type of conference and the number of users. The WWW conference client can be connected to the remote WWW conference client via WWW conference server to create the video stream. Our system uses the channel to create and manage the video stream. Each channel has a channel number. The channel number is



used for the clients connection and selection of the video stream. In the point to point conference and multi-user conference, the channel numbers are assigned based on the connection order. While, in the broadcast conference, the channel numbers are selected by the user. An example of the relation between the video stream and channel number in the point to point conference is shown in Figure 7. This conference uses two video streams in each WWW conference client. One is used for the sender client and another for the receiver client. The sender video stream and receiver video stream have different channel numbers. After allocating the channel number, the WWW conference client creates the video streams by using the channel number. In this example, the user 1 is connected to the WWW conference server before the user 2. For this reason, the channel number starts from 1. The channels 1 and 2 are shown in Figure 7. In the channel 1, the user 1 is the sender and the user 2 is the receiver of the video stream. While, in the channel 2, the user 2 is the sender and the user 1 is receiver. The relation between the video stream and channel number in the multi-user conference is shown in Figure 8. In this type, each user has its own video stream. The channel number is assigned the same as type one. After allocating the channel number, the WWW conference client creates the video stream for each receiver. In this example, four users are connected to WWW conference server. In the channel 1, the user 1 is the sender and the users 2, 3 and 4 are receivers. Each WWW conference client creates the video streams by using these channel numbers. The relation between the video stream and channel number in the broadcast conference is shown in Figure 9. The user can select the sender channel number using the sender client. Then, the video stream is created based on the selected channel number. Also, the user can select the receiver channel number using the receiver client. It

Implementation and Performance Evaluation of WWW Conference System

Figure 7. Connection management for point to point conference Sender Channel

R eceiver Channel

V ideo Stream

1

1 User 2

User 1

2

2

R eceiver Channel

Sender Channel

Figure 8. Connection management for multi-user conference Sender

Sender V ideo Stream Channel

R eceiver Channel

1

User 1

2

User 2

4

User 4

R eceiver

1

User 2

1

User 3

1

User 4

2

User 1

2

User 3

2

User 4

4

User 1

4

User 2

4

User 3

Figure 9. Connection management for broadcast conference Sender

Sender Channel

V ideo Stream

R eceiver Channel

1 1 User 1

User 2

1

2

n User n

1

R eceiver

User 2 User 3 User 4

2

User 3

n

User 1

n

User 4



Implementation and Performance Evaluation of WWW Conference System

should be noted that a user can not send the video stream in a channel when this channel is used by another user. But, the receivers can get the same channel at the same time. When the sender channel number has already been used, the WWW conference client displays this message: “This channel number has already been used.”

WWW conFErEncE oPEraTion The WWW conference operation has two phases: the start up phase and live video communication phase. In the start up phase, the WWW conference client is downloaded and the user enters into the WWW conference server using the WWW browser. The live video communication phase consists of camera setup, video quality modification, and sending and receiving the live video. The start up phase is shown in Figure 10. First, the user downloads the desired type of the WWW conference client from the WWW conference server. After that, the WWW conference client

starts up automatically and requires the user ID and password. When the user inputs the user ID and password, the WWW conference client sends this information to the WWW conference server to authenticate the user. When the authentication fails, the WWW conference client requests the user ID and password again. When the authentication is a success, the WWW conference client creates the video streams. Then, the WWW conference client connects to the shared object to share the name of recorded video stream with other WWW conference clients. The live video communication phase is shown in Figure 11. First, the WWW conference client asks the user for the permission to use the camera. Next, the WWW conference client allocates the channel number to send and receive the live video stream. When the WWW conference client finishes the channel allocation, it starts to send/receive the live video streams to/from other clients via WWW conference server. When the user requests to record the live video stream, the WWW conference client sends this request to the WWW conference server.

Figure 10. Startup phase WWW B r owser Download of W W W Conference Client

WWW Ser ver

R equest W W W Conference Client

W W W Conference Client

WWW C onfer ence C lient Start up for W W W Conference Client Input User ID and Password

R equest User A uthentication User A uthentication R esult of User A uthentication

WWW C onfer ence Ser ver

Connect to W W W Conference Server V ideo Stream Connect to Shared Object

R esult of Connect

0

Implementation and Performance Evaluation of WWW Conference System

Figure 11. Live video communication phase WWW C onfer ence C lient

WWW C onfer ence C lient

WWW C onfer ence C lient

-Permission for using camera -Prepare for receiving live video

V ideo and sound data V ideo and sound data -Permission for using camera -Prepare for receiving live video

-Start recording live video

WWW C onfer ence Ser ver

R equest record live video

V ideo and sound data

-Permission for using camera -Prepare for receiving live video

V ideo and sound data V ideo and sound data V ideo and sound data -R ecord live video

-R efresh video data name -R efresh video data name -R efresh video data name -Input frame size

V ideo data name Modify frame size -R efresh frame size

V ideo and sound data -Input receive frame rate

Modify receive frame rate -R efresh frame rate V ideo and sound data

The WWW conference server adds the new name of live video stream to the shared object. Then, the name of live video stream is added to the connected WWW conference clients. When the user inputs the frame size, the WWW conference client updates the frame size, modifies the captured frame size and sends the modified frame size to the WWW conference server. When the WWW conference server receives the modified frame size, it updates the new frame size and sends the live video stream to the WWW conference clients using the new frame size. Also, the frame rate is updated the same as the frame size.

imPlEmEnTaTion The implemented system structure is shown in Figure 12. The WWW conference client is implemented by using the Macromedia Flash Studio MX1.5. The WWW conference server is realized by using the Apache1.3.29 and the Mac-

romedia Flash Communication Server MX1.5. The Apache1.3.29 is used to provide the WWW conference client to the users. The Macromedia Flash Communication Server MX1.5 is used for sending and receiving the video data and voice data. The live video streams are encoded by H.263. We implemented three types of applications for each type of conference. When the user downloads the desired type of application, the WWW conference client displays the authentication interface. The user can take part in the WWW conference after authentication. The point to point conference application is shown in Figure 13. The bottom side of interface shows the remote live video, recorded live video, and local live video from left to the right side. The list box and play button are shown below the recorded video. The list box is used to select the name of video stream and play button is used to play the recorded live video stream. The stop button is shown in the left side below the local live video stream. This button is used to



Implementation and Performance Evaluation of WWW Conference System

Figure 12. Implemented system structure WWW Browser

WWW Conference Server WWW Server(Apache1.3.29)

User Interface

WWW Conference Client (Point to Point Version) • FLASH Application•

FLASH Player WWW Conference Client (Point to Point Version) • FLASH Application•

FLASH Communication Server MX1.5

WWW Conference Client (Multi-User Version) • FLASH Application• WWW Conference Client (Broadcast Sender Version) • FLASH Application• WWW Conference Client (Broadcast Receiver Version) • FLASH Application•

WWW Conference Client (Multi-User Version) • FLASH Application• WWW Conference Client (Broadcast Version Sender) • FLASH Application• WWW Conference Client (Broadcast Version Receiver) • FLASH Application•

Network Interface

Network Interface

Internet Message

Object

Video Stream

Figure 13. Point to point conference application

stop the recorded live video stream. The record button is shown in the right side below the local live video. The user can record and play the live video stream by using these buttons. The main video display can be selected by clicking one of

262

the videos shown in the bottom side. The video quality setup is shown in the upper side of the interface. The interface also has in the bottom side some list boxes, which are used to change the frame size and frame rate. The frame size can

Implementation and Performance Evaluation of WWW Conference System

Figure 14. Multi-user conference application

Figure 15(a). Broadcast conference sender application

Figure 15(b). Broadcast conference receiver application

be changed only at the local live video. But, the frame rate can be changed in both: the local live video and remote live video. The multi-user conference application is shown in Figure 14. In the bottom side of interface are shown the remote live videos and local live video. The right side displays the local live video. The user can modify the frame rate for each remote live video using the list box. Also, the user can modify the frame rate and frame size for the local live video using the list box.

The broadcast conference applications are shown in Figures 15(a) and (b). These applications include the channel select interface, video quality set up interface and video display interface. The sender client interface is shown in Figure 15(a) and the receiver client interface is shown in Figure 15(b). The main video in the center of the display shows the snap shot of the live video captured from the video camera. In the bottom side of interface are shown the text box, stop button and play button. The text box displays the name



Implementation and Performance Evaluation of WWW Conference System

of recorded video stream. On the right side of the interface are shown the sender channel select buttons. The user can select the channel number from 1 to 12. Also, the user can input the name of live video stream and record the live video on the WWW conference server. When the user pushes the channel select button, the sender client starts to send the live video stream and the name of live video stream using the selected channel number. The video quality set up interface has the list box shown below the video. The user can modify the frame size and frame rate by selecting the values in the list box. The video display interface displays the live video and interface select buttons. The receiver client interface is almost the same with sender client interface, but the sender client interface can record the live video while the receiver client interface can play the recoded video.

EXPErimEnTs Experimental results In order to evaluate our system for remote mental health care education, we used our system in a real Internet environment between Fukuoka, Ishikawa, and Iwate prefectures. Our evaluation environment is shown in Figure 16. The LANs are connected to the Internet environment by leased line (1 Gbps), FTTH (100 Mbps) and ADSL (12 Mbps). The experiment showed that the live video quality is good enough for communication. But it should be noted that sometimes the live video quality deteriorates because the bandwidth is not enough to send and receive the live video. In such cases, the frame rate is decrease and voice data is delayed about 1 second. Also, we evaluated the performance of our system on the LAN environment (100 Mbps) to analyze the possibility of realization of remote mental health care. We analyzed the throughput and load average for the

Figure 16. Evaluation environment Netbridge Company Head Office Takizawa, Iwate

Sunbeam-Hiyorigaoka Tatsuruhama-machi, Ishikawa WWW Conference Client

WWW Conference Client

ThinkPad X31 CPU: PentiumM(1.4GHz) Memory: 256MB OS: WindowsXP Camera: Logicool QV-4000

ATU-R

USB

Splitter

Router

ADSL 12Mbps

USB

VGA

ADSL 12Mbps

Leased Line 1Gbps

FireWall

RS232C

Modem

Internet FTTH 100Mbps

ONU

HUB

100Mbps

HUB

WWW Conference Client iBook G4 CPU: PowerPC G4(1GHz) Memory: 512MB HDD: 30GB OS: MacOS X 10.3.3 Camera: Logicool QV-4000



CPU: Pentium4(3.06GHz) Memory: 512MB HDD: 160GB OS: WindowsXP Camera: Sony EVI D30 Video Capture Card Brooktree Bt848

WWW Conference Server

CPU: Pentium3(1.7GHz) Memory: 512MB HDD: 80GB OS: Redhat8 Software: Apache 1.3.29 Flash Communication Server MX 1.5

Fukuoka Institute of Technology

Netbridge Company Morioka Office

Fukuoka-shi, Fukuoka

Morioka-shi, Iwate

Implementation and Performance Evaluation of WWW Conference System

Table 1. PCs characteristics for performance evaluation CPU

Memory

OS

Server

Pentium4 (1.6GHz)

1G

Windows XP

Client 1

Pentium3 (450MHz)

128M

Windows2000

Client 2

Pentium3 (667MHz)

128M

WindowsXP

Client 3

Pentium3 (664MHz)

128M

Windows2000

Cleint 4

Celeron (700MHz)

192M

WindowsXP

Figure 17. Relation between frame rate and throughput Number of client:2 Frame size:320x160 2500 2000 1500 1000 500 0

0

5

Server: Send(KB)

10

15 20 Frame rate[fps]

Server: Receive(KB)

Table 2. Relation between frame rate and load average Frame Rate 1 6 12 24 30 Load Average 40 69 87 93 94

multi-user conference. In this experiment, we use low performance computers because usual people use in general low performance computers. We used five PCs in this experiment. One PC is used as WWW conference server and four other ones are used as the WWW conference clients. The PC characteristics are shown in Table 1. From these results, we conclude that the bandwidth of the LAN environment is enough for our

25

Client: Send(KB)

30

35

Client: Receive(KB)

Table 3. Relation between frame size and load average Frame Size Load Average

160x120 320x240 640x480 30 69 92

system. But, the PCs should have more power to send, receive and display the live video. The relation between the frame rate and load average at client side is shown in Table 2 and the relation between the frame rate and throughput is shown in Figure 17. We see that the load average and throughput are affected by the frame rate at client side. But, we found that the throughput is not changed more than 12 fps. This is because



Implementation and Performance Evaluation of WWW Conference System

the PCs have not enough CPU power and the load average is close to 100%. After that, the throughput remains the same. In this case, the frame is delayed and the load average becomes high. The relation between the frame size and load average at client side is shown in Table 3 and the relation between the frame size and throughput is shown in Figure 18. We see that the load average is affected by the frame size at client side. But, we found that the throughput is not affected by frame size. The relation between the number of clients and load average at client side is shown in Table 4 and the relation between the number of clients and throughput is shown in Figure 19. We see that the load average and throughput are affected by the number of clients. Especially, the data sent from the server are affected by the number of clients because each client needs to receive all of live videos except for the own live video. But, the data sent from the client remain the same.

comparison of Proposed and conventional systems In Table 5 we compare our proposed system with conventional systems considering network environment, implemented platform, hardware/ software, installation and setup, and communication service. Some of conventional systems can be used not only in LAN environment/leased line, but also in Internet environment. However, these systems don’t run in the multi-platform OSs. Using these systems, each user needs to install and setup the special hardware or software. In some of them is needed only the setup, but each user needs to make installation of special hardware or software. Furthermore, there are few systems, which can be used for point to point, multi-point to multi-point, and broadcast communication. While, our proposed system can be used in real Internet environment, and Windows, Macintosh



and Linux OSs. In order to use our system, the user needs only to access the WWW conference server by using the WWW browser. After that, the system is installed and set up automatically. Also, the proposed system supports point to point, multi-point to multi-point and broadcast communication.

conclUsion In this chapter, we proposed a WWW conference system for supporting remote mental health care education. We showed the conference types and connection management methods for realizing each type of conference. We presented the WWW conference operation. Finally, we discussed the implementation and evaluation of our proposed system. From experimental results, we conclude as the follows: 1. The proposed system can provide the same communication between Fukuoka, Ishikawa, Iwate prefectures over the Internet. 2. The WWW conference client can send and receive the live videos for four users in the LAN environment. 3. The WWW conference client needs a high CPU power to send, receive, and display the live video.

Now, we are evaluating the proposed system for different kind of networks and number of hops. In the future, we would like to use the proposed system for mental health care education and counseling.

rEFErEncEs Abiko, T., Iijima, H., Koyama, A., Kamibayashi, N., & Narita, N. (2003). Implementation and verification of care communications service between hospitalized patients and their families

Implementation and Performance Evaluation of WWW Conference System

(IPSJ SIG Technical Report DPS114-6), pp. 3743. (in Japanese) Advanced Solutions. (2001). moNet. Retrieved from http://www. asi.co.jp/monet/index.html (in Japanese) Advanced Solutions. (2002). Impression live. Retrieved from http://www.asi.co.jp/imlive/index. html (in Japanese) Fuji-shi. (2005). Fuji-shi live camera. Retrieved from http://www.city.fuji.shizuoka.jp/live/ Hitachi Hybrid Network. (2004). IP visual communication system NetCS. Retrieved from http:// www.hitachi- hybrid.co.jp/business/net/netcs01. html Inoue, T., Okada, K., & Matsushita, Y. (1997). Spatial design for integration of face-to-face and video meetings: HERMES video conferencing system. IEICE, J80-D-II(9), 2482-2492. (in Japanese) Kato, V., Jiang, D., & Hakozaki, K. (2003). A proposal of a streaming video system in best-effort networks using adaptive QoS control rules. In Proceedings of IPSJ DPSWS’2003. IPSJ Symposium Series, 2003(19) (pp. 7-12). (in Japanese) Kawamura, N., Maita, Y., Hashimoto, K., & Shibata, Y. (2003). Remote medical support system using the transcoding function (IPSJ SIG Technical Report DPS 113-5), pp. 69-74. (in Japanese) Kumamoto-shi (2001). Today’s Kumamoto Castle. Retrieved from http://www.city.kumamoto.kumamoto.jp/ castleToday.html Macromedia. (2003). Flash Communication Server MX 1.5. Retrieved from http://www.macromedia.com/software/ flashcom/ Macromedia. (2004). Flash MX 2004. Retrieved from http://www.macromedia.com/software/ flash/

Maeda, K., Aibara, R., Kawamoto, K., Terauchi, M., Kohno, E., & Nishimura, K. (1997). Multimedia communication environment for distance learning. IEICE, J80-B-I(6), 348-354. (in Japanese) Microsoft. (2004). Netmeeting. Retrieved from http://www.microsoft.com/windows/netmeeting/ Miyoshi, K., Okanaga, Y., Kou, S., & Kondo, S. (2000). Design, development, and experiments of distance learning systems by satellite Internet services. IEICE, J83-D-I(6), 644-650. (in Japanese) Mori, K., Oyabu, Y., Nomura, A., & Oshita, S. (1992). Inter-campus tele-education and its evaluation. IEICE, J75-A(2), 244-255. (in Japanese) NEC Engineering. (1996). Remote conference system. Retrieved from http://www.nec-eng. com/kaigi/index.html (in Japanese) NTT West. (2004). Technical Report 12. Retrieved from http://www.ntt-west.co.jp/ipnet/ip/siryou. pdf (in Japanese) Polycom. (2004). Retrieved from http://www. polycom.com/home Sapporo-shi. (2004). Tokeidai at the moment. Retrieved from http://web.city.sapporo.jp/livecamera/ SONY. (2004). Video conference system PCS-1. Retrieved from http://bssc.sel.sony.com/Professional/markets/videoconferencing/overview. html Sakiyama, T., Ohono, N., Mukunoki, M., & Ikeda, K. (2001). Video stream selection according to lecture context in remote lecture. IEICE, 84-DII(2), 248-257. (in Japanese) The House of Councilors. (2005). The House of Councilors Internet TV. Retrieved from http:// www.webtv.sangiin.go.jp/webtv/index.php (in Japanese)



Implementation and Performance Evaluation of WWW Conference System

The House of Representatives. (1999). The House of Representatives Internet TV. Retrieved from http://www.shugiintv.go.jp/top.cfm The Ministry of Public Management. (2004). Home affairs, posts, and telecommunication. Retrieved from http://www.soumu.go.jp/media/ index.html Visual Nexus. (2004). Visual online conference system. Retrieved from http://www.visualnexus. com/en/index.html

VTEL. (2004). Retrieved from http://www.vtel. com/ Wakahara, T. (1998). Configuration and characteristics of distance learning system over ATMPVC Network. IEICE, J81-B-I(8), 494-506. (in Japanese). Yamaguchi, T., Sakano, T., Fujii, T., Ando, Y., & Kitamura, M. (2001). Design of medical teleconsultation support system using super high definition imaging system. IEICE, J84-D-II(6), 1203-1212. (in Japanese)

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 3, edited by S. Chang and T. K. Shih, pp. 77-96, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).





Chapter XVII

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning Degan Zhang University of Science and Technology of Beijing, China Yuan-chao Li China University of Petroleum, P.R. China Huaiyu Zhang Northwest University, China Xinshang Zhang Jidong Oilfield, P.R. China Guanping Zeng University of Science and Technology of Beijing, China

absTracT As a new kind of computing paradigm, pervasive computing will meet the requirements of human being that anybody maybe obtain services in anywhere and at anytime, task-oriented seamless migration is one of its applications. Apparently, the function of seamless mobility is suitable for mobile services, such as mobile Web-based learning. In this chapter, under the banner of seamless mobility, we propose a kind of approach supporting task-oriented mobile distance learning paradigm. Web-based seamless migration, which has the capability that task for mobile distance learning (MDL) dynamically follows the learner from place to place and machine to machine without learner’s awareness or intervention by active service. Our key idea is this capability can be achieved by architecture of component smart platform and agent-based migrating mechanism. In order to clarify the approach, firstly, a description Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

of the task for mobile distance learning and migrating granularity of task has been suggested. Then, the mechanism of seamless migration has been described, including solving several important sub-problems, such as transferring delay, transferring failure, residual computation dependency. Finally, our implemented platform for Web-based seamless migration has been explained, the validity comparison and evaluation of this kind of mobile distance learning paradigm is shown by an experimental demo. Suggested Web-based learning paradigm by seamless migration is convenient to distance learn during mobility and is useful for the busy or mobile distance learner.

inTroDUcTion It is known to all that pervasive/ubiquitous computing (Weiser, 1991) is a new computing paradigm fusing the technologies of computing, communication, and digital multimedia, which integrates information space and physical space of human being’s life, so it makes the computing and communication just like the life necessity, such as water, electricity, and air. This paradigm meets the requirements of human being that anybody maybe obtain services in anywhere and at anytime, so it is full of future. Nowadays, many ambitious projects have been proposed and carried on to welcome the advent of pervasive computing. There are a bunch of branch research fields under the banner of it, such as Seamless Mobility (Satyanarayanan, 2001). For seamless mobility, the history and context of computing task will be migrated with person’s mobility, and the computing device and software resource around this task will make adaptive change. The chief function requirement of seamless mobility is on the continuity and adaptability of computing task. The continuity is that the application can pause and continue the work without the loss of the current state and the running history. The adaptability is that the application is not restricted by computing device and context of service but adaptable to its environment. Apparently, this function of seamless mobility is suitable for mobile learning paradigm (Takasugi, 2001, 2003). For learner, it is necessary and accessible when he or she can NOT complete his or her learning task/courseware, such as video,

0

audio, text, picture, etc., in one specified scene, he or she can go on learning the uncompleted task/courseware in other spots by seamless mobility based on the Web. In our opinion, this is a kind of mobile working paradigm—learning by seamless migration with computing task. But when seamless migration for computing task of learning is realized on PC, laptop, or PDA, there are several difficult problems to be solved: (1) Meet different networked Web environment, such as different OS platform. (2) Manage the seamless-service among multiple machine devices. (3) Describe computing task of learning and only migrate the relative parts of task interested by learner in order to reduce the delay produced by migrated data. In this chapter, we propose a test bed of learning by seamless migration for mobile learning, which can be suitable for the required dynamic changes to the network and environment without learner awareness or intervention, and the condition of only sitting in front of the desktop PC for mobile learning is unnecessary. The structure, mechanism, result of experimental evaluation of the test bed is reported. It makes the ultimate mobile system possible by dynamically implementing the changes required to follow the learner from place to place and machine to machine. The rest of this chapter will be organized as follows. Firstly, we give formal description of task of mobile distance learning and migrating granularity of task of learning. After that, we design and discuss efficient approach of Seamless Mobility based on agent for task-oriented mobile distance learning, along with the description of

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

our implemented platform for Seamless Mobility. Finally, we evaluate the validity of the approach and platform for mobile distance learning and draw a conclusion.

DEscriPTion For TasK oF lEarning In order to clarify and realize how to transfer tasks of learning among different distance computing environments, firstly, a formal description and classification of task is required, which is independent of the realization mechanism. To adapt the environment of pervasive computing, a universal description language for task of learning should be used. Nowadays, the description languages for workflow or task of learning are mainly based on stationary computing environment (Simmons & Apfelbaum, 2001). However, the computing environment of seamless mobility is dynamic and mobile, so the description language should be abstract and self-adapted. Based on our knowledge, XML (extended markup language) and SMIL (synchronized multimedia integration language) released by W3C can be used (Shi & Xie, 2003). The task or transaction of learning cared by learner is our alleged Task (in brief, T), which consists of subtask or sub-transaction Ti, each Ti is an independent unit of function. Because of the diversity of task, its subtask or atomic task may be different from each other. In order to keep the compatibility, the description of subtask should be abstract, mainly, the key and necessary parameters, such as Qos of subtask, environments, etc. During mobile distance learning, the task can be classified into three kinds based on DATA TYPE of its contents: 1.

Event-Type: It is strict with the delay, the transferred bytes of subtask is few, but timely, semantic and no loss during trans-

2.

3.

ferring. Once the command of operation is done, the result should be shown. Stream-Type: It is not strict with delay and semantics, permitting a certain loss during transferring, but strict with jitteriness of transferring. Bulk-Type: It is different from event type because the transferred byte of subtask is much larger (maybe several Mbytes), and it is also different from stream type because when executed, it requires the integrality of data.

Described formally task of learning by SMIL is as follows: …….

mobilE granUlariTY oF ThE TasK How to deal with the problem of task-oriented mobile distance learning under the banner of seamless mobility? Currently, the usable technology is based on Mobile IP used as network-level protocol and stationary or mobile agent used in application-level. Active and intelligent mobile agent controlled by running container can deploy or adjust dynamically its services according to application requirement or running status of net-



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

work. As a kind of special computing resource, agent supports deployment of computing resource and mobility freely, which makes the system manage and adjust easily, so it is suitable for application of seamless mobility. During the task-oriented mobile distance learning, mobile granularity of the task should be traded off reliability, the communication volume of network, and so forth. According to integrity of transferred contents of the task, the mobile granularity of the task may be divided into “Strong Transfer” and “Weak Transfer,” and the mode of transferring may be controlled by the Travel Schedule/Plan. Strong Transferring means that total information involved in the current task must be transferred, after reaching the target terminal, the task can execute continuously from snapshot point. But in mobile WWW, it is difficult to collect total information of current task, to describe and record the executed status and necessity of task under the high bandwidth network, so the burden of this mode is very heavy and complex. Nowadays, the JVM is not supported this mode. Weak Transferring is only done for partial executing status and data, its speed is much faster then that of Strong Transferring, and its delay is much shorter then that of Strong Transferring. Of course, Weak Transferring has its shortcomings, for instance, the total historical executing status of task is difficult to be restored. So it is decided only by detailed scenario that which mode should be adopted in the application.

agEnT-basED aPProach oF TasK-oriEnTED sEamlEss mobiliTY Because the proliferation of mobile devices and the appearance of runtime environment give new challenges to support user mobility, we give a mechanism of integrating mobile devices into runtime environments to provide more compu-



tational, communication and storage capabilities to mobile distance learner. In this mechanism, we think that the data migrating is needed between different computing devices by different network, such as wireless infrastructure-based communication, multi-hop ad-hoc networks, dynamic topology without any infrastructurebased communication, Internet-based networks and different computing devices interconnected using IEEE 802.11x and Bluetooth technology. So a seamless and transparent migrating mechanism between different networking interfaces is needed. Seamless migrating between different networks for different computing devices by mobile agents is a basic feature for improving the quality of a perceived service under the pervasive computing mode. However, the heterogeneity also implies that the services are also distributed over the accessible networks. Based on context information, our mechanism spontaneously interoperates with available resources discovered in runtime environment so that it can improve the performance of mobile interactive distance learning.

Classification of Agent On the attribute of agent, it can be classified into two types. One is Stationary Agent, which is not mobile and kept in the agent environment (AE). The other is mobile agent (MA), which may be transferred in the AE. On the function of agent, it can be classified into three types, terminal agent (TA) (including user agent (UA)), navigation agent (VA) (including network capability agent (NCA) and location management agent (LMA)), task agent (KA) (including execution/code agent (EA) and data agent (DA), such as service agent (SA), user document database agent (UDDA))). These agents may be Stationary or Mobile Agent. Stationary or Mobile Agent (Danny & Mitsuru, 2001) is a kind of program with its name and can interact with other agents or resources when transferring from one network to another in different heterogeneous network (Karnik, 1991).

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

This program can dynamically decide when and where to transfer. It can suspend at any running point or transfer to another computer and execute continuously. Mobile Agent also can clone itself or produce its sub-agent and transfer to other computer to do complex task in cooperation mode. Besides common attributes (such as autonomous, initiative, intelligent), mobility is its main attribute, which makes it roam among different networks. Mobile Agent is suitable for computing of large heterogeneous network like Internet. Mobile Agent System (MAS) (Ciancarini, 2002) is a kind of system for creating, explaining, executing, scheduling, transferring, and terminating agent. Each system can run many agents. TA is one interactive interface for human and terminal devices. Users may search, browse, subscribe, and publish new service by TA. UA is on half of the user, which may transfer among different environments with the user. UA can partly store attribute data of the user and be used as buffer of current terminal. VA is used as addressing in the MAS and carry KA in its “MessageBox (MB).” KA is used as restoring runtime environment, executing task in the new environment, and recording each snapshot as current and history execution status.

new approach of seamless mobility based on agent The basic strategy of transferring based on agents has two modes: 1. 2.

Strong Transfer from source node to target node. Weak Transfer from source node to target node.

The first mode is adapted for seamless mobile scenarios with smaller amount of transferred data, such as task of “Event-Type.” If the task is “Stream-Type” or “Bulk-Type,” the delay of transferring and unnecessary amount of data are

larger, which is not adapted for seamless transferring on Mobile WWW. The second one may be implemented through two methods: partial information has been loaded on the target node/terminal station, downloaded the relative partial information timely during the runtime of task. This mode is adapted for the task with “Stream-Type” or “Bulk-Type,” which can reduce the transferred amount of data, the delay of transferring and improve the running efficiency, but occupied a certain storage space of target node. The basic transferring step of agents is as follows: 1. 2. 3.

4.

Determine the agents running on the source node. Suspend the agents. Snapshot or record the information of running agents and transfer them to target node. Reconstruct or restore the information of running agents on the target node.

In our opinion, the design of transferring method is considered from two aspects: 1. 2.

The transferring amount of data is total or partial. The occasion of suspending agent on the source node and restoring agent on the target node.

The existing transferring mechanisms (Takasugi, 2003) have NOT analyzed and discussed both of them deeply, especially, how to adapt for the application requirement of seamlessness of pervasive computing. In our opinion, it must deal with the problems of seamless transferring method, transferring delay, transferring failure and residual dependency, and so forth. Based on the basic strategy of transferring mentioned previously, now we discuss the new



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

efficient mechanism of seamless mobility suggested by us. If the transferred amount of data is partial, and this part of information must be transferred firstly so that the task can restore the runtime environment and run continuously on the target node, this part of information is named “Key Set,” such as executing code, running status, and so on. According to the classification of agent discussed earlier, we make the following rule: 1.

2.

3.

Navigation Agent (VA) need NOT do direct relative works with the task, which is familiar with the topological structure of subnet of target node and addressing in the network. The Data Structure of VA may be divided into two parts: one is itself “function body,” another is MessageBox (MB, mark as ) used as loading moved object and transferring in the subnet. Task Agent (KA) does detail jobs, which includes executing the code, managing the data and environmental status, and so on. It can transfer with the Navigation Agent (VA) in the network and need NOT know the subnet’s structure. When transferred, KA looks up relative VA and joins in its MB firstly, and then sent to target node by VA.

For the sake of convenience, we give a kind of general case: TA wants KA on the logic node PA 2 (Persona Avatar 2) to be transferred to another logic node PA3 (Persona Avatar 3), according to the timetopological relation of transferred object, the “TRAVEL SCHEDULE/PLAN” which is a kind of DATA STRUCTURE independent of agent has been made. The current scenario is that the TA is connected with logic node PA1 which is connected to logic node PA 2 through double direction link, The arrow shows the connected direction



and solid line with arrow shows KA can transit the logic link. The designed algorithm of seamless mobility includes eight main steps: 1.

2.

3.

4.

5.

According to the subscribed TRAVEL SCHEDULE/PLAN for transferring, logic node PA 2 lets VA begin addressing in the network according to the address supplied by logic node PA3, when the connection is successful, VA sends instruction “TransferNode” to Logic node PA3 as target node, VA+ transfer to PA3 after packing, the packet consists of the recorded structure of KA, the association relationship between KA, the space occupied by KA, the type of KA, the information of VA for task transferring and “Messenger” information for scheduling all agent (including VA, EA and DA), the state of arrived Messenger is not “Executing” but “Waiting” and storing in the queue of PA 3. Logic node PA 3 sends instruction “UpdateLinking” to all logic nodes connected to PA 2, such as logic node PA1 (Persona Avatar 1). The instruction includes the information modifying the link address, such as link ID, IP and Port of two ends. During the transferring, the Messengers to PA3 store in the relative queue and wait for executing unless the Key Set or the total task is finished to be transferred. When PA1 has received the instruction “UpdateLinking,” it creates the association to new link, and sends instruction “LinksUpdated” to logic node PA 2. When PA 2 has received all expected message “LinkUpdated,” and then sends instruction “ActiveNode” to logic node PA3. The message includes the list of all arrived Messenger. At the same time, PA 2 delete the transferred VA+. When PA3 has received the message “ActiveNode,” received Messenger from PA 2

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

6.

7.

8.

listed in the tail of the queue. According to the topological relationship, under the rule of FIFO, PA3 activates the Messenger. Up to now, the transferring work is finished. When all Messengers are activated, each KA will restore running environment and do instruction “ExecuteTask.” During the executing of each KA, on the one hand, the historical snapshots will be recorded and saved (including the structure of KA, the association relationship between KA, the space occupied by KA, the type of KA, the information of VA for task transferring and Messenger information for scheduling all agent (including VA, EA and DA), the state of Messenger for scheduling), on the other hand, VA do the instruction “ListenTask” continuously and get the next transferring instruction “TransferSignal.” During the executing of KA, if no instruction “TransferSignal” is got by VA, KA will execute its task continuously until the task is completed, otherwise, it will stop executing and go to Step 1 for preparing the new turn of transferring. The new turn process will be subject to the subscribed TRAVEL SCHEDULE/PLAN. The whole process of transferring is seamless.

agent1 arrives at node C, the agent2 can’t be found by the agent1. This case is so-called transferring failure problem. This kind of problem can’t keep the continuity of transferring of task. In order to solve this problem, there are three factors should be considered: 1.

2. 3.

When the position of agent has moved, how to know this change by other relative agents. When the transferring of agent, how to deal with the message sent to it. During the transferring of agent, whether the receiving agent can be transferred freely or not. Our solution is as follows:

1.

2.

3.

The previous approach is for a kind of general case. The other special cases is similar to this one, such as if TA wants to interact with PA2/PA3, PA1/PA2 should be involved according to the subscribed TRAVEL SCHEDULE/PLAN.

When the agent moves to new node, it should send “Notification” Message to all other relative agents. Before the agent prepares to be transferred, it should look up the current position of receiving agent, at the same time, the transferring relationship and event should be sent to it by message The transferring topological relationship of agent should be determined. We select the rule “FIFO (First In First Out)” for it. When some agents begin to be transferred, the other receiving agent should be locked. After the transferring process is over, the receiving agent should be unlocked at once. Based on this rule, a signal semaphore may be set up.

Transferring Failure Problem In the pervasive computing environment, because the position of agent is often variable, the cases may be occurred. When the agent1 is being transferred to agent2 and wanted to be embedded in agent2 on node C to deal with the task together, but the agent2 has moved from node C to node D during the transferring of the agent1. That is to say, when

The “Notification” Message may adopt three kinds: Unicast, Multicast, Broadcast. For different applications, in detail, it can be divided as follows: 1.

Unicast, Multicast, Broadcast in the GROUP



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

2. 3. 4.

Unicast, Multicast, Broadcast among the GROUPS Broadcast from a agent to MAS Unicast, Multicast, Broadcast among different MAS

The formal expression of “Notification” Message of agent is that: Agent_Message_Notify Message)

(Sender,

Receiver

,

Correspondingly, the formal expression of Unicast, Multicast, Broadcast is that: Agent_Message_Notify (agent1, agent2, Message) Agent_Message_Notify (agent, Multicast (agentX) , Message) where “agentX” shows other agents Agent_Message_Notify (agent, Broadcast (Any), Message) where “Any” shows any agent of Groups Based on the “time-topological” relationship, we design a kind of synchronal mechanism “addressing first, then locking and transmitting,” which can realize the synchronization between transferring agent and receiving agent. So the transferring failure problem can be solved radically and adapted for all kinds of application pattern. The “time-topological” relationship can be used in the “Travel Schedule for transferring.” The schedule may consist of certain travel sequences, each travel sequence includes the following DATA STRUCTURE: Schedule ID TP_ID, Schedule Name TP_Name, Schedule Made Date TP_Time, Travel Number No, Transferring Object TP_Object, Transferring granularity TP_Granule, Source Address of Transferring Object TO_IPC, Target Address of Transferring Object TO_IPD, Transferring Condition



TP_Condi, Transferring Mark TP_SnapshotPoint, Entry Address for Re-running TP_RunEntry. Transferring Condition TP_Condi, Transferring Mark TP_SnapshotPoint, Entry Address for Re-running TP_RunEntry are important for basic transferring operation, that is to say, Only the TP_Condi is OK, the “Travel Schedule for transferring” may be run, meanwhile, record and save “TP_ Snapshot Point” and “TP_RunEntry,” both for restoring the running environment. Whether TP_Condi is OK or NOT, the following aspects should be set and checked: 1.

2. 3. 4.

Whether or not the current status (may be divided into five kinds: Ready 1, Waiting 2, Transferring 3, Running 4, Destroyed 5) of agent is “Waiting 2.” Whether the target address TO_IPD may be reached or not. Whether the threshold of transferring delay is OK or not Whether the residual dependency cases may occur or not.

Transferring Delay It is named “Transferring Delay” that the time interval from the suspended snapshot point of a running agent to the re-run snapshot point on the target node. The delay is one of main parameters to access the seamless mobility. The information for transferring an agent includes that instruction sets, address sets, runtime state when suspending, executing code, data, Messenger schedule information, and so on. Messenger schedule information and runtime state after being suspended must be transferred totally, but other information may not be transferred totally. Once the necessary information has been restored on the target node, especially the snapshot point of runtime state before being transferred, the agent may be re-run.

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

Three factors is mainly involved in the transferring delay of agent: 1. 2.

3.

Which kind of transferring granularity, Strong Transfer or Weak Transfer. When will the agent be transferred, which is the suspending time of the agent. The occasion of suspending agent is divided into three kinds: Suspend immediately after determining to transfer, Suspend after the total information is transferred completely, Suspend after the Key Set is transferred successfully. When will the agent be restored and run, which is the restoring time of agent. The occasion of restoring agent on the target node is divided into two cases: restore after the total information is transferred completely, restore after the Key Set is transferred successfully.

In fact, there are three valid mapping forms based on the previous three factors: 1. Strong Transfer: Suspend after transferring, restore and run after finishing the whole information. 2. Strong Transfer: Suspend after transferring, restore and run timely after finishing the Key Set information (at the same time, the whole residual information will continue transmitting). 3. Weak Transfer: Suspend after transferring, restore and run after finishing the Key Set information (at the same time, the selected partially information will continue transmitting to the target node). In the first mode, the transferring delay of agent includes that packing the whole information and transmitting, restoring the agent and the whole information on the target node. In the second mode, includes that packing the Key Set information and transmitting, restoring the agent and the Key Set. In the third mode, includes that

packing the Key Set information and transmitting, restoring the agent and the Key Set. In the same condition, the delay of the first one is longest, the third one in shortest.

residual computation Dependency Problem In the second and third transferring mode, because of selecting a part information as the “Key Set” and being transferred firstly, but different applications, it is NOT known which part of information is necessary. So a certain “Key Set” may NOT transferred to the target node timely, the running agent must wait for it. That is to say, running agent is still dependent on part of information on the source node. This case is named “residual computation dependency.” This problem may lengthen the transferring delay of agent, when serious, it will influence the seamlessness. So the cases must be avoided. In our opinion, the “residual dependency” problem will be solved from two aspects: 1.

2.

Tuning reasonably the transferring granularity of Execution/code Agent (EA), Data Agent (DA) and other agents (such as environment-state agent). If too larger, it is restricted by the bandwidth. If too smaller, transferring time is much more. Both may lengthen the delay. Based on analyzing from theory and application tests, we suggests a partition method named “subsection” or “pagination,” which determines the size or number of “section” or “page” by bandwidth, buffer, volume of MessageBox. When the cases are occurred, the necessary information may be transmitted through “section interruption” or “page interruption,” but the frequency should be adjusted automatically according to historical record information. Optimize the Key Set. The Key Set will be determined automatically according to near-



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

est principle and used frequently principle and cut off the redundancy information. The relative adapted strategy may be referred the bibliography (Milojicic, 2002; Takasugi, 2001).

communication Primary for migration In order to support Web-based seamless migration, we have designed several communication primaries for transferring agent, a part of them are as follows: 1.

BeginToListen Primary: VA will invoke the BeginToListen Primary to listen the port nPort, meanwhile, register the callback function OnAccept to receive connection message from other agents, when VA have received the connection request, the OnAccept will call back the connection_ID Connection_ ID. The Primary is: BeginToListen(UINT nPort,ACCEPT_ CALLBACK callback); Void (CAgent::* ACCEPT_CALLBACK) (UINT& Connection_ID). BeginToRequest Primary: When VA wants to set up connection to target agent, it will invoke BeginToRequest Primary to send request to the agent with IP:nPort, if it is successful, the connection ID nConnection_ID will be called back. The Primary is: BeginToRequest(UINT &nConnection_ID, CString IP,UINT nPort). Tranfer Primary: When agent wants to send messages or transfer task to other objects, it will invoke Tranfer Primary to do it. The Primary is: Transfer (UINT nConnection_ID, CString strMsg, CDate time_stamp).

2.

3.

Similarly:



PrepareForRecv Primary for receive message or data stream: PrepareForRecv(UINT nConnection_ID, RECEIVE_CALLBACK callback); Void (CAgent::* RECEIVE_CALLBACK) (CString &strMsg, UINT& ConnectionID). PrepareForClose Primary for close or end the connection: PrepareForClose(UINT nConnection_ID, CLOSE_CALLBACK callback); Void (CAgent::* CLOSE_CALLBACK) (UINT &Connection_ID).

ThE imPlEmEnTED PlaTForm The function and service of software platform (Simon, 2002) supporting task-oriented mobile distance learning paradigm—Web-based seamless migration should include: 1.

2.

Management method of resource and services: When a new mobile device is brought into a space or new module or component used in the old device, the software manager can know how to spontaneously discovery them and what is wanted to be interactive. Because the resources of device are not same in a system, they may be embedded device, wearable computing device, mobile computing device, etc. their computing capability, memory capability, interactive mode are different. When the device is mobile or nomadic in the different environment, the interconnection problem is existed. The infrastructure can transform or translate the contents. Support for message-oriented, streamoriented or bulk-oriented communication: Here we argue that actually there are three catalogs of communications needs in runtime environment of mobile distance learning with different QoS requirements.

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

Message-oriented communication is a kind of communication that occasionally happens and usually has high-level semantics for distance learning, e.g. a command asking a module or component to play the specified information. These communications are sensitive to the loss of messages; whereas their requirements on the delivery latency are moderate, as long as it is within a reasonable boundary, say, 50 ms, according to the cognitive character of human. Stream-oriented communication is a kind of communication that constantly occurs. Their semantic level usually is relatively low and the drop of several data units is usually tolerable. However, they are sensitive to the variation of the delivery latency, while in most cases their requirement on the delivery latency is also moderate. Bulk-oriented communication is a kind of communication with much larger Bytes amount (maybe several K/Mbytes) to be delivered, which

3.

is not much sensitive to the variation of the delivery latency. The work paradigm may be Client/Server, Browse/Server and Peer-toPeer, so the protocol stack of communication is a certain kind of link of IP—TCP/UDP/ RTP—HTTP/FTP. Coordination mechanism of continuity and self-adaptability: As a distributed mode, the infrastructure can coordinate the relationship of association, communication, collaboration of modules, so coordination mechanism of continuity and self-adaptability among modules or component is more important to the whole function and services. Of course it is important that supporting one-to-many communication, heterogeneous platforms and implementation languages. It is common in runtime environment of mobile distance learning that a message should be delivered to many modules or components simultaneously. Even in the case of stream-oriented com-

Figure 1. Structure description of platform A p p lica tio n /A g e n t S m art P latform A P I/X M L

S tream

S ys S ervice

M essage

B ulk

C ontrol li nk w ith A synchronous m essage h andle R T P /U D P IP M ulticast E nvironm ent D iscovery M odule

TCP

X M L/T C P IrD A

B luetooth P roxy A gent

E thernet/W LA N S ystem A P I O S /N etw ork



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

Figure 2. Structure of seamless migration embedded in the platform

Se am les s M ig ratio n P lat form OS ND

A

S M -se ssio n la ye r

T

S Mco n n e ctio n la ye r

T

A

S M -p a th la ye r

T

A C

TC P /IP la ye r

T

MA

A

T

MA

A

T

C

S M -p a th

S M -p a th

S M -lin k

S M -lin k(IP ,IP )

IP A d d re ss M A C A d d re ss

L in k la ye r W ire d L a n

munication, there will be multiple learners of a single stream. Therefore, it is necessary for software platform to have the one-tomany communication capability. Modules or components in a runtime environment usually impose different requirements on the underlying hardware and OS. Moreover, they are often implemented in different languages. A software platform should have the adequate capability to deal with all these diversities. Based on analyzing to the function and service of software platform, we have designed and developed it. Figure 1 is the structure description of our implemented platform supporting task-oriented mobile distance learning paradigm — Web-based seamless migration. Figure 2 is the structure of seamless migration embedded in this implemented platform, which includes four layers: SM-link layer, SM-path layer, SM-connection layer and SM-session layer. In Figure 2, “T” stands for “Task,” ”A” stands for “Agent,” “MA” stands for “Mobile Agent” and “C” stands for “Container,” which is a daemon threads component installed in each relative mobile devices. Their working principle has been previously mentioned.

0

A

S M -co n n e ctio n

S M -lin k la ye r

P h ysica l la ye r

MA S M -se ssio n

W ire le ss L a n

W AN

Our implemented platform can work in Client/Server, Browse/Server and Peer-to-Peer paradigm. The platform is a multi-agent system. The structure can be divided into multiple levels. Multiple agents are collaborated for seamless mobility. Each one has its special function. The agent class and container class can be partially defined as follows: class CAgent { public: CAgent(); virtual ~CAgent(); BOOL Register(); BOOL Quit(); BOOL Subscribe(CString strGrpName, NOTIFY_ CALLBACK callback, CString strTemplate=””); UINT GetSharedFile(LPCTSTR url,LPCTSTR lpszTagInfo=NULL); virtual void OnConnect(); virtual void OnDisconnect(); ... }; class Ccontainer { public: CContainer(); virtual ~CContainer(); BOOL LaunchAgentByName(CString strAgtName); BOOL LaunchAgentByPath(CString strPath);

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

void ProcessDSCmd(CDSMsg & msg); …; typedef struct _MINIHTTP_REQUEST { SOCKET socket; char* http_data; unsigned long http_data_size; MINIHTTP_FIRST_LINE* first_line; } MINIHTTP_REQUEST; typedef struct _MINIHTTP_RESPONSE { unsigned int range_begin; unsigned int range_end; unsigned long http_data_size; int http_response_code; } MINIHTTP_RESPONSE; ... }

There are four kinds of components in our platform: Management Interface Component, Task Manager Component, Continuity Manager Component, and Service Manager Component. The interface is often used as defining the attributes of agent, such as ID of Agent, Name of Agent, Type of Agent (such as TA, SA, UA, VA, EA, DA, and so on), Current Status of Agent (five status are “Ready,” “Waiting,” “Transferring,” “Running,” “Dead” or “Destroyed”), Association Relationship of agent (including relationship between agent and task, relationship between two agents). The task manager is for application service, which manages the application/task array, including task description, task analyzing, mapping, or binding between task and service, loading, executing, scheduling of task, etc. The continuity manager is for agent management, context-awareness computing, and history/status recording, such as making of “Transferring Travel Schedule” of agent, addressing of VA, determining of transferring granularity which is for avoiding the transferring failure, reducing the residual dependency and contracting the transferring delay, etc. The service manager conducts the registration of service, discovery of service, service association, and mapping or binding between task and service of mobile distance learning. Service discovery is the base of seamless mobility of task of learning. Currently, several discovery ideas have been designed or used, such as Service Location

Protocol, Jini, Salutation, Universal Plug and Play, Bluetooth Service Discovery Protocol, and others (Garlan & Siewiorek, 2002). These components can communication each other, and may be controlled by application interactive interface including agents and global control of task, which is interface of human computers, such as PC, laptop, PDA, Mobile phone, embedded devices. The stationary or mobile agent is the basic encapsulation of the software modules in the system for management of service and mobility. Each computer in the runtime environment of mobile distance learning will host a dedicated process called Container, which provides system-level services for the agents that run on the computer and manages them as well. It makes the details of other parts of the system transparent to agent and provides a simple communication interface for stationary or mobile agent. There is one global dedicated process in the environment, which mediates the “delegated communication” between stationary or mobile agent and provides services such as directory service, dependency resolution.

TEsT oF mobilE DisTancE lEarning basED on oUr PlaTForm Our implemented platform can show many scenarios, such as Web-based seamless migration with “Event-Type” task, “Bulk-Type” task, “Stream-Type” task for task-oriented mobile distance learning. Here is an example that includes Web-based seamless migration for task of learning on PC, laptop, or PDA under dynamic changes of the network and environment without user awareness or intervention. Just like Figure 3, the task of learning can follow me form one device to another device or from my house to other places, such as my office, stadium, coffee house, park, airport, etc., and vice versa.



Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

Figure 3. Task can be migrated with me from one place to another place

Office

Coffee house

Now supposing the task of learning consists of three sub-tasks of mobile distance learning: playing video, playing mp3, and reading documents. As a demo of many scenarios, this kind of task of learning is described partially by SIML as follows:



House

Stadium

Park

Airport

The previous task of learning has three subtasks of learning: Playing Pervasiv.avi, playing IloveChina.mp3, reading Pervasive.txt. They will be done according to the time-sequence in parallel mode based on the algorithm mentioned previously. With the learner’s movement from one station (such as House) to another station (such as Airport), these uncompleted sub-tasks of distance learning can seamlessly migrate from PC of his or her house to laptop, or PDA with him or her in Airport and go on learning (watching, listening, and reading) continuously by mobile agent on our platform supporting Web-based seamless migration. In our experiments of mobile distance learning, the deployment of device is the CPU frequency, RAM of PC and laptop (a kind of mobile device) are 1.2 GHz, 512 MBytes,respectively, and 450 MHz, 64 MBytes RAM are for PDA (another kind of mobile device), the speed of wired network and wireless network is 10 M/100 MHZ, 1-3 MHZ, respectively. The nodes were connected by wireless and wired Web network—Internet.

Web-Based Seamless Migration for Task-Oriented Mobile Distance Learning

conclUsion In order to meet the application requirements of mobile distance learning, we have proposed a kind of novel distance learning paradigm — taskoriented Web-based seamless migration, which supplies the function that the task of distance learning dynamically follows the learner from place to place and machine to machine, so it is convenient to learn during mobility, and is useful or helpful for the mobile learner or mobile attendee. Our key idea is that this capability can be achieved by layering architecture of component platform and agent-based migrating mechanism. In this chapter, we have given the formal description of the task of mobile distance learning, discussed the migrating granularity of the task of learning. The innovative significance is that we have designed a kind of approach of Web-based seamless migration, including solving these problems, such as shortening migration delay, avoiding migration failure and residual computation dependency. The validity of this approach and its corresponding software platform for mobile distance learning has been tested by many demos.

rEFErEncEs Ciancarini, P. (2002). Coordinating multi-agent applications on the WWW: A reference architecture. IEEE Trans. on Software Engineering, 24(5), 363-375. Danny, B. L., & Mitsuru, O. (2001). Seven good reasons for mobile agents. Communications of the ACM, 42(3), 86-89. David, K., & Robert, S. G. (2002). Mobile agents and the future of the Internet. ACM Operating Systems Review, 33(3), 7-13.

Garlan, D., & Siewiorek, D. P. (2002). Project aura: Toward distraction-free pervasive computing. IEEE Pervasive Computing, 1(2), 22-31. Karnik, N. M. (1991). Design Issues in mobile agent programming systems. IEEE Concurrency, 6(3), 125-132. Milojicic, D. (2002). Mobile agent applications. IEEE Concurrency, 7(3), 80-90. Satyanarayanan, M. (2001). Pervasive computing: Vision and challenges. IEEE Personal Communications, 8(8), 10-17. Shi, Y. C., & Xie, W. K. (2003). The smart classroom: Merging technologies for seamless teleeducation. IEEE Pervasive Computing Magazine, 2(2), 25-33. Simmons, R., & Apfelbaum, D. (2001). A task description language for robot control. In Proceedings Conference on Intelligent Robotics and Systems, New York (Vol. 1, No. 10, pp. 138-147). Simon, S. (2002). A model for software configuration in ubiquitous computing environments. In Proceedings of Pervasive. LNCS 2414, Zürich (Vol. 1, No. 7, pp. 181-194). Takasugi, K. (2001). Adaptive system for service continuity in a mobile environment. In Proceedings of IEEE APCC, Tokyo, Japan (Vol. 1, No. 9, pp. 75-83). Takasugi, K. (2003). Seamless service platform for following a user’s movement in a dynamic network environment. In Proceedings of PerCom’03 (Vol. 1, No. 8, pp. 125-132). Weiser, M. (1991). The computer for the twentyfirst century. Scientific American, 265(3), 94104.

This work was previously published in International Journal of Distance Education Technologies, Vol. 4, Issue 3, edited by S. Chang and T. K. Shih, pp. 62-76, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).





Chapter XVIII

Digital Rights Management Implemented by RDF Graph Approach Jin Tan Yang Southern Taiwan University of Technology, Taiwan Huai-Chien Horng National Kaohsiung Normal University, Taiwan

absTracT This chapter proposes a design framework for constructing digital rights management (DRM) that enables learning objects in legal usage. The central theme of this framework is that any design of a DRM must have theories as foundations to make the maintenance, extension or interoperability easy. While a learning objective consists of learning resources and its metadata, a DRM also needs metadata for describing itself as rights expression language (REL). The proposed resource description framework (RDF) graph design in this study is based on the Boolean operations of graph theory, whereas the RDF graph provides not only more coherent operations, but also opportunities for maintenance and interoperability at different platforms. Two algorithms for encoding and verifying rights in DRM are designed to deal with REL metadata in RDF format. This technological support also reduces the sophistication among role assignments, learning objects and task ontology of DRM. The DRM module is embedded to SCORM-compliant content repository management system (CRMS) for IPR (intellectual property rights) protection. Finally, some implications of this study are also included.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Digital Rights Management Implemented by RDF Graph Approach

inTroDUcTion The digital revolution, powered by the engines of information and communication technologies (ICT), has fundamentally changed the way people think, behave, communicate, work and earn a living. It has restructured the means by which the world conducts economic and business activities and runs governments. Moreover, it has formed new ways to create knowledge, educate people and disseminate information. Recently, the MIT Open CourseWare (OCW) has offered more than 700 free of charge courses, including lecture notes, course materials, examinations and lecture videos (OCW, 2004). Thus, the OCW can be accessed by anyone eager to learn domain knowledge in the global village, free from any physical boundary. Although MIT has been a pioneering institution in learning technologies field, the real potential problem is to keep content providers who are willing to provide high quality teaching materials and to protect their contents in legal usage. In this digital era, the digital rights management (DRM) not only refreshes e-learning content, but also leads us towards a new model of education (Rosenblatt, Trippe & Mooney, 2001). However, the enactment of DRM for content providers in education must rely on better algorithms to deal with Digital Rights metadata. ICT has been adopted for access control in terms of protecting content providers rights. Many approaches have been proposed and applied in the real market such as Microsoft, Adobe or IBM. Although those leading companies have digital rights to make their rewards, the rewards only apply to specified customers, data formats or delivering platforms. Thus, we must develop a new way for learning objects to be protected by access control. Generally speaking, most of the existing access control mechanisms implemented for Web applications can be classified into two major categories: role-based access control (RBAC) models and the hypertext-based authorization models (Lu & Chen, 2003).

One of the core components of RBAC models is a role that represents different organizational responsibilities and functions. The use of this role can simplify the task of authorization administration by organizing related access privileges to a role and assign users to the role (Sandhu, Coynek, Feinsteink & Youmank, 1996). Later, if a user is promoted to a new position in the organization, the user can simply be assigned to a new role and removed from the old one. In addition, RBAC models support various security policies such as role hierarchies and constraint. Traditionally, the objects under control of RBAC models are either programs or documents. In this study, we extend it to sharable content object reference model (SCORM), compliant learning objects or content packages (ADL, 2004). Originally, a role in RBAC can access rights of the whole course package but, not portions of the course package. This feature, however, seriously limits its applicability on the Web. To overcome it, we adopt the resource description framework (RDF) graph with Boolean operations to unite different learning objects as a unit. A RDF graph consists of nodes and arcs (Brickley & Guha, 2000). Nodes are labeled either with an URI (concept Resource) or an atomic value (concept Literal). Thus, a role with access rights, encoded as RDF graph, on a course package in our study can access a content package or a portion of the package. Content repository management system (CRMS), a collection of SCORM-compliant learning objects, has been developed (Yang & Tsai, 2003). It, however, did not consider DRM while it authorizes a system administrator as the gate keeper for assignments of digital rights of learning resource. It violates content providers as primary authorization to confer rights for any potential users. In this study, we assume that only content providers can decide their contents to users with permissions. Summarily, this study proposes a mechanism to enact the DRM for content providers within



Digital Rights Management Implemented by RDF Graph Approach

legal usage. To describe the authorization language, RDF graph is selected as rights expression language (REL) language instead of traditional XML-based language such as ODRL (open digital rights language) (ODRL, 2003; XrML, 2002) or XACML (extensible access control markup language) (Guth, Neumann , & Strembeck, 2003). The reason for the RDF graph being chosen is that RDF graph has superiority to deal with the complexity of rights assignment, and Boolean operation for learning objects or content packages. Moreover, REL in RDF file formats can be used to reason by graph matching. In this study, content providers encode the digital rights in RDF file. Then, the system will decode digital rights from RDF file while users access those learning objects or content packages in CRMS. With support of semantic RDF graph, this study proposes a simpler, machine processsble and extensible model for DRM. In the following section, related literature on DRM will be reviewed and our solution will be proposed. Followed by literature review, the functionalities of RDF graph with Boolean operations will be investigated and the implementations be presented. After the discussion and conclusion, future studies will be recommended in the last section.

where we are, it still has inclusive definitions so far. The Association of American Publishers points out that DRM consists of the technologies, tools and processes that protect intellectual property rights (IPR) during digital content commerce, whereas some (Fetscherin & Schmid, 2003; Iannella, 2001; Rosenblatt, 2001) state DRM offers intellectual property (IP) Asset Creation/Capture, Management and Usage in terms of functionality. In other words, there are two purposes in DRM. One is that using ICT to protect IPR for rewarding the content providers. The other one that both content providers and consumers can use legally by DRM system reduces the conflict or chaos between content producers and content consumers. Practically, DRM has three major components (Downes, Babin, Belliveau, Blanchard, Levy, Bernard, Paquette & Plourde, 2004): 1.

2.

3.

liTEraTUrE rEViEW This chapter proposes a design framework for constructing DRM that enables learning objects in legal usage. To reach the goal, some foundational theories such as DRM, ontology and RDF graph with Boolean operations may offer the rationale.

Digital right management (Drm) Although digital right management (DRM) issue has been recognized as high ranking in cyberspace



Expression: To describe the resource, ownership of the resource and the terms and conditions of use such as rights expression language (REL), closed security policy, agreements or contracts. Authentication: To verify that using the resource meets the rights association to the resource such as Graph-based validation and Rule-based validation. Protection: To ensure only authorized users who are able to access by such mechanisms such as encryption, for example, copy detection systems, digital signature and information security systems. This kind system often serves as infrastructure layer.

In this study, we focus on proposing a REL and a graph-based verification as rights enforcement of REL. Protection of information security technologies, however, is beyond the scope of this chapter, it can be extended by REL through its extensibility.

Digital Rights Management Implemented by RDF Graph Approach

The Definition of Right Expression language (rEl) Right expression language (REL) consists of three basic elements: rights, assets and parties (Guth, 2003). First, rights are represented as expressions which grant certain usage or access permissions to digital goods or services. Compared to rights, permissions can be a more specified or restricted asset. Third, the party element might represent a legal role or physical person with unique ID, such as a student with an ID “Jim.” For example, a consumer, called “Jim,” is allowed to view math assets at junior high levels with permission, but only three times. Such kinds of propositions should be described by a set of vocabularies or REL. Once the knowledge of a domain is represented in a declarative formalism, the set of objects that can be represented is called the universe of discourse. This set of objects, and the describable relationships among them, are reflected in the representational vocabulary with which a knowledge-based program represents knowledge (Gruber, 1993). To represent the real context in DRM, many REL, such as ODRL (2002), VRML (2001) and so forth, have been proposed and supported by different organizations. Traditionally, REL uses XML as representation language, such as ODRL, XrML, XACML and so forth. The XML data model is a text-markup oriented labeled tree. And as XML and XML schema are designed primarily for fixed, tree-like documents, they are significantly inflexible for expressing metadata of rights expression, which by its very nature is subjective, distributed and expressed in diverse forms. RDF, by contrast, has a very simple model consisting of labeled arcs that forms graph-like data and is also simpler, more flexible to meet REL’s requirements. Furthermore, any specific set of RDF statements forms a graph that can be serialized in XML and inherit its benefits. In terms of rights verification, it is inclusive in both research and industry field. Traditionally,

most systems use rule-based method and some others use closed encoded program to implement the inference of verification (Guth, Neumann & Strembeck, 2003). Those approaches make interoperability across applications and platforms difficult. To provide more coherent and interoperable operations, RDF graph verification model adopted graph theory and its operations to refine the task. This approach benefits mainly from shift partial inference logics to data-layer. Thus, this study adopts RDF graph as descriptor of REL to enhance its expression power by graph approach.

rDF graph/ontology Ontology is based on RDF/RDFS in terms of development stages (Berners-Lee, Hendler, & Lassila, 2001). Thus, we discuss RDF first. The foundation of RDF is based upon graph theory since it is a directly labeled graph in RDF/RFS representation (Hayes,2003). To explain it, RDF/RDFS consists of three components such as subject, predicate and object in sequential order. For example, “This Thesis” (subject) is “author” (predicate) and “Horng, H.C.” (object) in Figure 1. In RDF/RDFS expression, the above sentence can be expressed as follows: Horng, H. C. Second, ontology is an explicit specification of a conceptualization (Gruber, 1993). In the context of AI, scholars describe the ontology as a program by defining a set of representational



Digital Rights Management Implemented by RDF Graph Approach

terms. Such definitions associate the names of entities in the universe of discourse (e.g., classes, relations, functions or other objects) with humanreadable text describing what the names mean and formal axioms that constrain the interpretation and well-formed use of these terms. Therefore, ontology can be regarded as the statement of a logical theory that the first-order logic deals with propositions (Grau, 2004). Subject and predicate in propositions are separately signified, reasoning whose validity depends on the level of articulation, and systems containing such propositions and reasoning. Guarino and Giaretta (1998) suggest that a conceptualization should be a triple structure , where W is the set of possible words, D is the domain of discourse, W is max state set in D. For example, the domain D is (S1, S2, S3, S4, S5), which are divided into two piles, (S1, S2, S3) and (S4, S5). Each pile is a state. When we put S1 on S4, we will get a new state. The collection of all states is called the max state set, W. R is the set of all conceptual relations in , n dimensional conceptual relations ρn is full function:

ρn: W 2 Dn To sum up, DRM must have theories as foundations to make the maintenance, extension or interoperability easy. RDF/ontology can be used to represent and infer among a set of vocabularies or propositions. Especially, the link between RDF graph and first-order logic makes inference feasible.

Figure 1. A sample of RDF model



access control in iPr Protection In traditional fields of access control system, there are three models: discretionary access control (DAC), mandatory access control (MAC) and rolebased access control (RBAC) (Lu & Chen, 2003). The major difference between DAC and MAC model is that content providers own entire rights to any user, whereas organization administrators in MAC model own content rights to any user. RBAC is always used to reduce the complexity of rights assignment between users and documents. The roles of users can be updated while their duty or positions are changed. Moreover, RBAC has minimum permission, duty discrepancy and permission inheritance. Thus, it offers an efficient access control in modern computer operating system. In analysis of the RBAC model for learning object in CRMS, RBAC is composed of three kernel parts: Users (U), Roles (R), and Permission (P) as shown in Figure 2 (Sandhu et al., 1996). Permissions can be divided into two components: “Resource” and “Operations.” The former can be mapped to learning object such as SCO or Asset. The latter can be mapped to operations such as “view,” “download,” “author” and so forth. Moreover, the link between permission and role in RBAC model is made by Permission Assignment (PA). In the same vein, User Assignment (UA) connects relations between “User” and “Role.” RBAC model is suitable for information systems in an enterprise in which an employee has a unique role in an organization. Thus, system administrator can assign permissions for individuals. It, however, is not the same story as CRMS while there is no predefined role for individuals. In other words, any learning resource cannot preassign permission for individuals. Therefore, content providers must confer rights of learning resources for individuals. Table 1 shows the comparison of RBAC model between enterprise and CRMS.

Digital Rights Management Implemented by RDF Graph Approach

Figure 2. RBAC Model adapted from Sandhu, et al.(1996).

Table 1. Comparison on RBAC model between enterprise and CRMS Items Ownership Asset authorization Authorized subjects

DRM in enterprise Department or Enterprise Administrator Precise organization framework

Therefore, DAC model for CRMS must be added into RBAC while content providers own entire rights to any user. Meanwhile, each individual’s role can be classified by teacher/student, discipline or system administrator to meet the requirements by RBAC model when an individual registers on CRMS.

rDF and graph Theory In graph theory, a graph is composed by vertices and edges, denoted G = (V,E) (Diestel, 2000). There are three basic Boolean set operations: union, intersection and difference. RDF graph, a derivation of graph theory, can also perform Boolean operations in terms of graph mapping. More precisely, RDF model can be used to construct different ontologies as graphs, and thus operations can be conducted as merging and disjointing among these ontologies occur (Jannink, 1999). In a RDF graph for semantic networks, the rules and facts can easily be fed into a template interpreter. Inference engines, based on selected, linear and definite (SLD) in logic programming, can be built while the structure of RDF has the triple sets by making certain relationships are explicit.

DRM on CRMS Content provider Content provider Everyone

Let us examine the graph with Boolean operations on union, intersection and difference in Figure 3 and Figure 4. Let G = (V, E) and G' = (V',E') be two distinct Graphs. Where

V = {v | ∀v ∈ G} ∧ E = {e | ∀e ∈ G}, V' = {v ' | ∀v ' ∈ G '} ∧ E ' = {e ' | ∀e ' ∈ G '} Figure 4 shows the results of three Boolean operations between G and G’ visually. The logical notations are shown in (1), (2), and (3).

G ∪ G': = (V ∪ V', E ∪ E') G ∩ G': = (V ∩ V', E ∩ E') G − G': = (V − V ∩ V', E − E ∩ E')

(1) (2) (3)

mEThoDologY This chapter proposes a design framework for constructing DRM that enables learning objects in legal usage. Two new approaches proposed in this study are: RDF graph REL representation and RDF graph verification. The former is given to



Digital Rights Management Implemented by RDF Graph Approach

Figure 3. Original graph: G and G’ (Diestel, 2000)

Figure 4. Graph representations of Boolean operations: (1), (2), and (3) (Diestel, 2000)

Figure 5. Mapping CRMS resources into RBAC model

0

Digital Rights Management Implemented by RDF Graph Approach

provide more flexible and semantic ways to describe digital rights of learning resource. The latter offers a simplified verification mechanism.

rights Expression mechanism RBAC model provides an easy way to clarify basic rights elements (Lu & Chen, 2003; Sandhu et al., 1996) by using roles, users and permissions. Figure 5 shows three elements and their relationship. For example, Resources-Operation pairs form permission rights. On the one hand, the resource consists of objects such as Asset, SCO or content package. Yet, operations include view, download or author play. The combination between resource and operation might be “Content Packages-Download.” The roles can be assigned by content providers, user’s domain and so forth. Such role classification is helpful in defining vocabularies in REL representation. In terms of task ontology (Fensel, Motta, Decker, & Zdrahal, 1997; Ikeda, Seta, Kakusho, & Mizoguchi, 1998), the role classification can

be further divided as shown in Table 2 with instances of each classification. For example, user contribution has “normal user,” “creator” and so forth. Once content providers decide to release rights to everyone, they can set the role classification as others.

Data structure as rDFs graph RDF schema (Brickley & Guha, 2000) is a mechanism that lets developers define a particular vocabulary for RDF data (such as “hasWritten”) and specify the kinds of objects to which these attributes can be applied (such as “Writer”). RDF schema does this by prespecifying some terminology, such as Class, subClassOf and Property, which can then be used in application-specific schemata. RDF schema expressions are also valid RDF expressions. In fact, the only difference with normal RDF expressions is that an agreement is made on the semantics of certain terms in RDF schema and thus on the interpretation of certain statements. For example, the subClassOf property

Table 2. Role classification and instances Role classification User Contribution User Identification Users’ Discipline System Level Others

Role instances {[Normal User], [Creator]…etc} { [Teacher], [Student]…etc} { [Chinese], [Math], [Science]…etc} { [Member], [Internal Staff],[Administrator]…etc} { [Everyone]}

Figure 6. An example of relationship between RDFS and RDF



Digital Rights Management Implemented by RDF Graph Approach

allows the developer to specify the hierarchical organization of classes. Objects can be declared to be instances of these classes using the type property. Constraints on the use of properties can be specified using domain and range constructs (Broekstra, Kampman, & van Harmelen, 2003). An example of “Writer” is illustrated in Figure 6. In this study, RDF schema is used to encode REL ontology as machine-interpretable form for system operation. The REL ontology shown in Figure 7 consists of fives entities: Asset, Party, Permission, Constraint and Requirement. The definitions of those entities are given as follows: 1.

Party might include individuals or rights holders with unique Party identification. 3.

Permissions are the actual usages or activities of assets and content packages. In CRMS context, the Permissions include View, Download, Author and Play. 4.

Constraint: Attribute Set = {usage count between 1 and n}

Constraints are limitations to access Permissions. For example, usage count constraint means the maximum number of times to access these assets.

Asset: Attribute Sets: Asset_UID and Asset_Type = {content package, Asset}

The Assets, a primitive learning object in SCORM definition, include any physical or digital content. Specifically, CRMS has two types of Assets, such as SCORM-compliant content package and SCO. Those Assets must be uniquely identified. Any new type of Assets can be easily added or deleted through these two attributes. 2.

Permissions: Attribute Sets = {View, Download, Author, etc.}

5.

Requirement: Requirements are the obligations needed to access the Permissions.

The REL ontology graph in Figure 7 can be encoded as RDF Schema in machine-readable format as shown in Table 3.

Party: Attribute Sets = Party_UID, and Party_Type set = {Role, User}

Figure 7. REL ontology graph based for CRMS Constraint Constraint

Asset isRightsof hasConstraint

hasRights

isConstraintof

hasPermission Permission

Rights isPermissionof

ownsRights hasRequirement

Party

isRequirementof

isOwnedby Requirement Requir



e-

Digital Rights Management Implemented by RDF Graph Approach

Table 3. REL ontology described by RDFS (Partial) ]>



Digital Rights Management Implemented by RDF Graph Approach

Figure 8. An example of encoding propositions in RDF graph

Table 4. Algorithm: Generation of usage rights graph Input : “User Request{User ID, Asset, Service}” Output : “Usage Rights Graph (URGr)” Declare URGr as RDF_Graph FOR each requested(LO) URG .CreateStatement(Asset, hasname, requested(LO)) END FOR URG .CreateStatement(Right, haspermission, requested(permission)) URG .Createstatement(Party, role, requested(user_name)) RETURN URGr

Table 5. An algorithm: RIP generates LRGr by Union operation Input : “Rights Instance Base” Output : “Learning-object Right Graph (LRGr)” Declare LRGr, LRGr’ as RDF_Graph(s) READ all Rights_Instance_RDF from LO_base as LRGr’ IF requested(LO).quantity > 0 FOR each LRGr LRGr = Union(LRGr, LRGr’) END FOR ELSE LRGr = LRGr’ END IF RETURN LRGr



Digital Rights Management Implemented by RDF Graph Approach

a scenario on Encoding rights Propositions in rDF graph Based on the RDFS graph in the above section, each proposition can be represented as a RDF statement. Thus, it gives an intuitive semantic of precise thinking. For example, Teacher A has rights to view the learning object in condition “c” and “r” whereas Teacher B has rights to download on the Learning Object in Figure 8. The propositions for Teacher A and Teacher B can be organized to represent rules and ontologies as shown in Figure 8. The proposition for Teacher A is “hasRights” (Teacher A, Rights isRightsof “LO” and hasPermission, “View”).

Rights Verification Mechanism The verification engine is used as a reasoning engine and is consulted before every usage request or taking any action. Based on the answer of the verification engine, user’s requests might be allowed or rejected. So a verification engine plays an important role by ensuring rights execution in DRM. The main design concept of rights verification mechanism is checking whether there are any inconsistencies between rights usage re-

quest graph (URGr) and learning-objects rights graph (LRGr). By transforming usage request information into graphs, it can be compared with graphs of authorized rights information through Boolean operations of graphs. Access is allowed if two graphs are matched, otherwise denied. The process of the reasoning engine is shown in Table 5. In Figure 9, RDF instance generator (RIG) gets information from users. The request from users can be represented as URGr. Thus, rights validation module (RVM) can retrieve rights of learning object by rights instance processor (RIP) and return a LRGr. In this section, Right Validator and RVM will be discussed in detail. First, the algorithm for Right Validator is shown as follows: Input : (1) “User request{User ID, Asset, Service}” (2) “Rights Instance Base” Output : (1) “Access granted or denied” (2) “Insufficient right information when denied.” Second, RVM is a bridge between LRGr and RIP. Thus, Urgr and LRGr must be designed before RVM mechanism.

Figure 9. Framework of rights validator



Digital Rights Management Implemented by RDF Graph Approach

Figure 10. RIG generates URGr

Figure 11. An example of generating LRGr by RIP

URGr: An algorithm in Table 4 can be explained how a URGr can be created by RIG in Figure 9. Single LO and multiple LOs requests are explained in Figure 10 respectively. RIG generates the two URGr in both sides of Figure 10. The former is that a teacher calls for viewing learning object #1, while the latter calls for two learning objects simultaneously. The algorithm of URGr is shown in Table 4. LRGr: While most users will take many learning objects at one time, to verify the permission, multiple learning objects from RDF file should be united before sending it to RVM. The algorithm of union operation is shown in Table 5. In other words, LRGr’ = LRGr1 È LRGr2. An example for union operation is shown in Figure 11.



Finally, the main design concept of rights verification mechanism is checking whether there are any inconsistencies between rights ontology graphs. Thus, RVM is in charge of matching between LRGr and URGr by RDF graph. To get the difference, the difference is applied. The algorithm is represented in Table 6. Assume IRGr = URGr – LRGr. If IRGr = ∫ then “Access Granted” Else “Access Denied” For example, Teacher A owns view rights on learning object #1 in  of Figure 12. Also, the

Digital Rights Management Implemented by RDF Graph Approach

Table 6. Algorithm: Difference of URGr to LRGr Input: “URGr”, “LRGr” Output: “Result of Validation”, “Insufficient Rights Graph (IRGr)” Declare IRGr as RDF_Graph REQUEST URGr, LRGr IRG = Difference(URGr, LRGr) CASE IRGr is

∫ :

RETURN messege(“Access Granted”) CASE IRGr is not

∫ :

RETURN messege(“Access Denied”) & IRGr ENDCASE

Figure 12. Access right is granted: Difference between URGr and LRGr is empty

Figure 13. Access right is denied: Difference between URGr and LRGr is notempty



Digital Rights Management Implemented by RDF Graph Approach

LRGr of learning object #1 is constructed in  of Figure 12. The result of matching  and  is  of Figure 12. In this case, the access right is granted while IRGr is empty. In the same vein, if difference between URGr and LRGr is not empty, then the access right is rejected as shown in Figure 13.

imPlEmEnTaTion

software Tools: Jena aPi

Encoding of learning objects in rDF Format

Jena is a Java API which can be used to create and manipulate RDF graphs. It was also originally developed at HP Labs Semantic Web Program. In this study, the Boolean operations of RDF graphs are implemented by calling Jena API methods. Thus, this study can deal with only system level, not finer levels such as coding in RDF implementation.

This section will demonstrate functionality by scenarios. The scenarios presented here are encoding of learning objects in RDF format on CRMS and a verification process while users try to obtain learning objects.

CRMS has a portal for users to log in as shown in Figure 14. Assume a content provider, called “jimhorng,” authorizes a user, named “Koach,” a learning object with rights: view and download. After the user follows the procedures to upload his/her SCORMTM learning object, the user’s graphic interface turns to Figure 15. “Jimhorng”

Figure 14. CRMS portal

Figure 15. The GUI authorized encoding on CRMS 1

2



3

4

Digital Rights Management Implemented by RDF Graph Approach

Figure 16. Rights assignment of Figure 15 encoded by Jena in RDF format - - limit View to 3 times - View Download

2 3

- - jimhorng_0_50_13 -

1

koach

4

is asked to fill the five slots in Figure 15. Most of slots are menu-driven for the user to choose from instead of inputting. For example, the user assigns an asset, called “jimhorng_15_3” to Kaoch by rights constraint consisting of view and download. The system will automatically generate an unique ID, called “jimhorng_0_50_17” for the content. The RDF file is encoded in Figure 16. Once Koach wants to retrieve “jimhorng_15_3” the RDF file will be parsed and generate as LRGr as shown in the above section. If URGr from Koach is matched to LRGr, then Koach can view or download the learning object. The GUI of Figure 15 is also allowed jimhorng to change the user authorization to individuals or group.

Verification Mechanism CRMS offers a function to allow users to search for learning objects or content packages by menudriven keywords or intelligent agents. In Figure 17, Koach has found 10 learning objects for assembling the user’s content package. While the user chooses learning object #10, three functionalities such as author, view and download are given to Koach. If the user clicks on “author”, then the user will be rejected in doing so. Conversely, the user can do it to the others because LRGr and URGr in RDF graph are matched. While Koach tries to view the learning objects, the user has been noticed that the system

299

Digital Rights Management Implemented by RDF Graph Approach

Figure 17. Three functionalilities: author, view, and download on CRMS

Figure 18. Access right is granted

Figure 19. Request for author

Figure 20. Access right is denied

00

Digital Rights Management Implemented by RDF Graph Approach

is validating his/her request. Later on, Figure 18 shows that Koach has been granted to view the learning object. Conversely, if Koach tries to author the learning object in Figure 19 and the user will be rejected while the content provider did not authorize him/her to author online in Figure 20.

Comparisons Between RDF Graph and Rule-Based Approach for DRM In traditional artificial intelligence, most scholars choose XML as REL and adopt rule-based approach to verify mechanisms in DRM. This study proposes a new approach by RDF as REL and RDF graph approach to verify DRM mechanisms of SCORM-compliant learning objects or content package. The RDF approach is superior to rulebased approach with at least three advantages, which are summed up as follows: 1. Simplification complexity of DRM: RDF as REL to represent DRM is based on graph theory and first-order logic. Thus, a programmer can encode and verify in RDF approach simpler than of rule-based approach. Actually, a RDF graph contains a set of propositions as ontology consists of a set of rules for verification. Through Boolean operation, both encoding and verifying are simplified. Moreover, the separation between encoding and decoding processes in this study is to reduce the complexity of DRM tasks. 2. Machine interoperability in RDF graphs: RDF graphs consist of a set of rules. It means that those rule logics are also put in data layer. Thus, a RDF API program can parse RDF files to construct RDF graph. In other words, a computer program just needs to do Boolean operations of RDF graph instead, to check all rules encoded at rule-based programs. Moreover, RDF files can be shifted to different platforms or programming languages while RDF user has

a standard API. Conversely, the rule-based approach is tightly coupled with specified platform or programming language. 3. Adaptation to different contexts: While new rules are added to RDF file or graph, there is no any change in inference engine. It reduces the burden of maintenance of computer system.

Discussion Learning objects or content packages access control decisions are still always identified by RBAC. Such access rights to learning objects is granted or rejected according to user’s personal information, such as account name and password. Once they enter to leaning content management system (LCMS) or LMS (leaning management system), they can navigate, search, view or download any learning object they want. It keeps content providers from supplying their learning objects to content repository continuously if they feel out of control on their content packages or SCOs. Therefore, digital right management should be assured in finer levels, such as who owns authorization to access a learning object. Indeed, each content provider should own the entire rights to assign their contents that are to be shared with specified group or individuals. Authorization schemes should be given by content providers while learning objects or content packages are uploaded. Verification schemes of accessing a content package should be also examined. The RDF as REL has been adopted as task ontology for encoding and verifying users’ rights. If a user requests a learning object, the digital rights of learning object by content providers will be generated; the comparison of two digital rights between content providers and users’ access right is compared with Boolean operations in terms of RDF graph. A user is granted or rejected to access a learning object or content package based on whether RDF graphs given by content provider and users’ request is matched. 301

Digital Rights Management Implemented by RDF Graph Approach

conclUsion We are certainly facing opportunities to share and to collaborate with unimaginably huge digital content repositories from now on. Many educational content repositories will encounter the digital rights issue if we expect or encourage content providers to develop more high-quality learning objects in the future. This study has proposed a resolution to coping with the complexity of DRM in terms of RDF graph. This chapter so far has focused on the design of DRM at CRMS. Certainly, a sound algorithm design by graph theories is critical for maintenance, interoperability in a LCMS. However, the ultimate goal of this study is to provide meaningful authorization for content providers and to verify the users’ requests for specified learning objects or content packages. Therefore, the evaluation of DRM at CRMS is essential for future research. In the near future, we will design a field assessment for performance and user satisfactions of RDF graph approach for DRM. Finally, the success of technology-based DRM does not solely rely on technology. Considerations of human cognition, and the social context of that thinking, take precedence over technology. To this aim, people must be conscience of digital rights in a real social context.

rEFErEncEs ADL. (2004). Sharable Content Object Reference Model (SCORM) 2004 (2nd ed.). Retrieved July 27, 2006, from http://www.adlnet.org/index. cfm?fuseaction=SCORMDown Berners-Lee, T., Hendler, J., & Lassila, O. (2001). The semantic Web. Scientific American, 284(5), 34-43. Brickley, D., & Guha, R. (2000). Resource Description Framework (RDF) schema specification

0

1.0 recommendation. World Wide Web Consortium. Retrieved July 27, 2006, from http://www. w3.org/TR/2000/CR-rdf-schema-20000327 Broekstra, J., Kampman, A., & van Harmelen, F. (2003). Sesame: An architecture for storing and querying RDF data and schema information. In D. Fensel, J. Hendler, H. Lieberman, & W. Wahlster (Eds.), Semantics for the WWW. Diestel, R. (2000).Graph theory (2nd ed.). Germany: Springer. Downes, S., Babin, G., Belliveau, L., Blanchard, R., Levy, G., Bernard, P., Paquette, G., & Plourde, S. (2003, April 22–23). Distributed digital rights management: The EduSource approach to DRM. In Proceedings of the First International ODRL Workshop, Vienna, Austria. Fensel, D., Motta, E., Decker, S., & Zdrahal, Z. (1997). Using ontologies for defining tasks, problem-solving methods and their mappings. In E. Plaza et al. (Eds.), Knowledge acquisition, modeling and management. Springer-Verlag. Lecture Notes in Artificial Intelligence (LNAI) 1319. Fetscherin, M., & Schmid, M. (2003). Comparing the usage of digital rights management systems in the music, film, and print industry. In Proceedings of the ICEC 2003. Pittsburgh, Pennsylvania. Grau, B. C. (2004). A possible simplification of the semantic Web architecture. In Proceedings of the WWW2004. New York. Gruber, T. R. (1993). A translation approach to portable ontologies. Knowledge Acquisition, 5(2), 199-220. Guarino, N. (1998). Formal ontology and information systems. In Proceedings of FOIS ‘98, Trento, Italy. Guth, S. (2003). Rights expression languages. In E. Becker, W. Buhse, D. Günnewig, & N. Rump (Eds.), Digital rights management: Technological, economic, legal and political aspects (pp.

Digital Rights Management Implemented by RDF Graph Approach

101-112). Heidelberg, Germany: Springer. Guth, S., Neumann,G., & Strembeck, M.(2003). Experiences with the enforcement of access rights extracted from ODRL-based digital contracts. In Proceedings of the ACM Workshop on Digital Rights Management, Washington, DC. Hayes, P. (2003). RDF semantics. World Wide Web Consortium. Retrieved July 27, 2006, from http:// www.w3.org/TR/2003/PR-rdf-mt-20031215/ Iannella, R. (2001). Digital Rights Management (DRM) architectures. D-Lib Magazine, 7(6). Ikeda, M., Seta, K., Kakusho, O., & Mizoguchi, R. (1998). Task ontology: Ontology for building conceptual problem solving models. In Proceedings of the 13th European Conference on Artificial Intelligence ECAI’98. Jannink, J. (1999, July). Ontology maintenance with an algebraic methodology: A case study. In

Proceedings of 1999 AAAI Workshop on Ontology Management, Orlando, Florida. Lu, J. L., & Chen, R. F. (2003). Design and implementation of a fine-grained menu control processor for Web-based information systems. Future Generation Computer Systems, pp. 1105-1119. OCW. (2004). The MIT OpenCourseWare update: July 2004. Retrieved July 27, 2006, from http://ocw.mit.edu/OcwWeb/Global/AboutOCW/ newsletter.htm ODRL. (2003). Open Digital Rights Language. Retrieved July 27, 2006, from http://odrl.net/ Rosenblatt, B., Trippe, B., & Mooney, S. (2001). Digital rights management: Business and technology (1st ed.). New York: John Wiley & Sons. Sandhu, R. S., Coynek, E. J., Feinsteink, H. L., & Youmank, C. E. (1996). Role-based access control model. IEEE Computer, 29(2), 38-47. XrML. (2002). Extensible Rights Markup Language specification. Retrieved July 27, 2006, from

This work was previously published inInternational Journal of Distance Education Technologies, Vol. 4, Issue 4, edited by S.K. Chang and T. K. Shih, pp. 36-55, copyright 2006 by IGI Publishing, formerly known as Idea Group Publishing (an imprint of IGI Global).

0

304

Compilation of References

A Skills Cert programme. (n.d.). The Irish Computer Society. Retrieved from http://www.ics-skills.ie/hompage.aspx Abiko, T., Iijima, H., Koyama, A., Kamibayashi, N., & Narita, N. (2003). Implementation and verification of care communications service between hospitalized patients and their families (IPSJ SIG Technical Report DPS114-6), pp. 37-43. (in Japanese). Abowd, G.D. (1999). Classroom 2000: An experiment with the instrumentation of a living educational environment. IBM System Journal, Special Issue on Pervasive Computing, 38(4), 508-530. Abowd, G.D., Harvel, L.D., & Brotherton, J.A. (2000). Building a digital library of captured education experiences. In Proceedings of the 2000 International Conference on Digital Libraries: Research and Practice, Kyoto (pp. 467-474). ADL. (2004). Sharable Content Object Reference Model (SCORM) 2004 (2nd ed.). Retrieved July 27, 2006, from http://www.adlnet.org/index.cfm?fuseaction=SC ORMDown ADL. (2004, January). SCORM run-time environment (RTE), version 1.3. Adomavicius, G., & Tuzhilin, A. (2001). Expert-driven validation of rule-based user models in personalization applications. In Proceedings of the 5th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD-99), New York, NY. Advanced Distributed Learning (ADL). (2003). SCORM present. Retrieved from http://www.adlnet.org

Advanced Solutions. (2001). moNet. Retrieved from http://www. asi.co.jp/monet/index.html (in Japanese) Advanced Solutions. (2002). Impression live. Retrieved from http://www.asi.co.jp/imlive/index.html (in Japanese) Aho, A.V., Hopcroft, J.E., & Ullman, J.D. (1983). Data structures and algorithms. Reading, MA: AddisonWesley. Ahrens, J. D., & Sankar, C. S. (1993). Tailoring database training for end users. MIS Quarterly, 17(4), 419-439. AIMS: Adaptive Information System for Management of Learning Content. (2004). Retrieved from http://wwwis. win.tue.nl:8080/AIMS Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179-211. Alavi, M. (1994). Computer-mediated collaborative learning: An empirical evaluation. MIS Quarterly, 18(2), 159-174. Allison, C., Ruddle, A., McKechan, D., & Michaelson, R. (2001). The architecture of a framework for building distributed learning environments. In Proceedings of the 2001 IEEE International Conference on Advanced Learning Technologies (pp. 29-32). Wisconsin: IEEE Press. Amoretti, M., Bertolazzi, R., Reggiani, M., Zanichelli, F., & Conte, G. (2006). Designing Grid services for multimedia streaming in an e-learning environment. Concurrency and Computation: Practice and Experience, 18(8), 911-923..

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Compilation of References

An 8-stage process to CPD for IT professionals. (n.d.). The British Computer Society (BCS). Retrieved from http://www.bcs.org/server.php?show=nav.1

Ayersman, D. J., & Minden, A. (1995). Individual differences, computers and instruction. Computers in Human Behaviour, 11(3-4), 371-390.

Andersen, T. (2004). Toward a theory of online learning. In T. Andersen (Ed.), Theory and Practice of Online Learning. Athabasca University.

Azjen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl & J. Beckmann (Eds.), Action control: From cognition to behavior (pp. 11-39). New York: Springer-Verlag.

Anderson, J.R. (1980). Cognitive psychology and its implications. San Francisco: Freeman. Aniebonam, M. C. (2000, October). Effective distance learning methods as a curriculum delivery tool in diverse university environments: The case of traditional vs. historically black colleges and universities. Communications of the Association for Information Systems, 4(8), 1-35. Apple, D.K., Nygren K.P., Williams, M.W., & Litynski, D.M. (2002). Distinguishing and elevating levels of learning in engineering and technology instruction. In Proceedings of 32nd ASEE/IEEE Frontiers in Education Conference, (pp. 6-9). Boston. Ardissono, L., & Cohen, R. (1995, August). On the value of user modeling for improving plan recognition. In Proceedings of the Workshop on The Next Generation of Plan Recognition Systems: Challenges for and Insight from Related Areas of AI (pp. 8-12). Arndt, T., Chang, S. K., Guercio, A., & Maresca, P. (2002). An XML-based approach to multimedia software engineering for distance learning. In Proceedings of the 14th International Conference on Software Engineering and Knowledge Engineering, (pp. 525-532). Ischia, Italy. Aslanian, C. (2001). Adult students today. New York: The College Board. Cited in Paradigm, 13(3). Retrieved March 2006, from http://www.mercyhurst.edu/graduate/paradigm-newsletter-pdf/fall2001.pdf Atkinson, S. (1998). Cognitive style in the context of design and technology work. Educational Psychology, 18(2), 183-194. Aviation Industry CBT Committee (AICC). (n.d.). Retrieved from http://www.aicc.org

Baeza-Yates, R., & Ribeiro-Neto, B. (1999). Modern information retrieval. New York: Addison Wesley. Baldwin, T.T., & Ford, J.K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41, 63-105. Baloian, N., Middleton, C., Breuer, H., & Luther, W. (2005). Algorithm visualization using concept keyboards. In Proceedings of the ACM Symposium on Software Visualization (SoftVis ’05), St. Louis, MO, USA (pp. 7-16). Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. Bandura, A. (1977). Social learning theory. Morristown, NJ: General Learning Press. Bandura, A. (1997). Self-efficacy: The exercise of selfcontrol. New York: W.H. Freeman and Company. Barbe, W., & Milone, M. (1980). Modality. Instructor, 89(6), 44-46. Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128, 612-637. Bauer, M. (1994). Quantitative modeling of user preferences for plan recognition. In Proceedings of the 4th International Conference on User Modeling (pp. 73-78). Bauer, M. (1996). Machine learning for user modeling and plan recognition. In V. Moustakis, & J. Herrmann (Eds.), Proceedings of the Icml’96 Workshop. Machine Learning Meets Human Computer Interaction (pp. 5-16). Baumgartner, P. (1999). Evaluation of media-based learning (in German). In M. Kindt (Ed.), Projektevaluation in

0

Compilation of References

der Lehre – Multimedia an Hochschulen zeigt Profil(e) (pp. 61-97). Münster: Waxmann. Bayman, P., & Mayer, R. E. (1988). Using conceptual models to teach BASIC computer programming. Journal of Educational Psychology, 80(3), 291-298. Bell, M.A., & Jackson, D. (1992). Visual author languages for computer-aided learning. In Proceedings of IEEE Workshop on Visual Languages, (pp. 258-260). Seattle, WA, USA.

Bostrom, R. P., Olfman, L., & Sein, M. K. (1990). The importance of learning style in end-user training. MIS Quarterly, 14(1), 101-109. Bouras, C., & Gkamas, A. (2003). Multimedia transmission with adaptive QoS based on real time protocols. International Journal of Communications Systems, 16, 225-248. Boyce, T. E., & Hineline, P. N. (2002). Interteaching: A strategy for enhancing the user-friendliness of behavioral arrangements in the college classroom. The Behavior Analyst, 25, 215-226.

Benetazzo, L., Bertocco, M., Ferraris, F., Ferrero, A., Offelli, C., Parvis, M., et al. (2000). A Web-based distributed virtual educational laboratory. IEEE Transactions on Instrumentation and Measurement, 49(2), 349-356.

Bradshaw, D. (2006, December 11). An MBA is what you make of IT. Financial Times, pp. 14

Bentley, J., Weide, B., & Yao, A. (1980). Optimal expectedtime algorithms for closest point problem. ACM Transactions on Mathematical Software, 6(4), 563-580.

Braune, B., & Wilhelm, R. (2000). Focusing in algorithm explanation. IEEE Transactions on Visualization and Computer Graphics, 6(1), 1-7.

Berners-Lee, T., Hendler, J., & Lassila, O. (2001). The semantic Web. Scientific American, 284(5), 34-43.

Brickley, D., & Guha, R. (2000). Resource Description Framework (RDF) schema specification 1.0 recommendation. World Wide Web Consortium. Retrieved July 27, 2006, from http://www.w3.org/TR/2000/CRrdf-schema-20000327

Bielefield, A., & Cheeseman, L. (1997). Technology and copyright law. New York: Neal-Schuman Publishers, Inc. Biermann, H., & Cole, R. (1999). Comic strips for algorithm visualization (Tech. Rep. 1999-778). New York University. Biggs, J. B., & Moore, P. J. (1993). The process of learning (3rd ed.). Englewood Cliffs, NJ: Prentice Hall. Biggs, J.B., & Collis, K.F. (1982). Evaluating the quality of learning – the SOLO taxonomy. New York: Academic Press. Blackboard Inc. (n.d.). Retrieved from http://www. blackboard.com Bloom, M. (2003). E-learning in Canada: Findings from 2003 e-survey. Bolot, J., Crépin, H., & Garcia, A. (1995). Analysis of audio packet loss in the Internet. In Proceedings of the 1995 Workshop on Network and Operating System Support for Audio and Video (pp. 163 -174). INRIA, France.

0

Britain, S., & Liber, O. (2004). A framework for pedagogical evaluation of virtual learning environments (Technical Report). Bolton Institute for Higher Education. Broekstra, J., Kampman, A., & van Harmelen, F. (2003). Sesame: An architecture for storing and querying RDF data and schema information. In D. Fensel, J. Hendler, H. Lieberman, & W. Wahlster (Eds.), Semantics for the WWW. Brotherton, J.A., Bhalodia, J.R., & Abowd, G.D. (1998). Automated capture, integration, and visualization of multiple media streams. In Proceedings of the IEEE International Conference on Multimedia Computing and Systems (pp. 54-63). Brown, M. H. (1988). Perspectives on algorithm animation. In Proceedings of the ACM SIGCHI ’88 Conference on Human Factors in Computing Systems (pp. 33-38). Washington, DC: ACM.

Compilation of References

Brown, M. H., & Hershberger, J. (1998). Program auralization. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software visualization: Programming as a multi-media experience (pp. 137-144). Cambridge, MA: MIT Press.

CAE SimuFlite. (2003, January 22). CAE to develop seLearning™ courseware for ALSTOM power plants (Press Release). Retrieved February 18, 2003, from http://www.cae.com/en/newsroom/2003/shtml/power_ 01222003_ref066.shtml

Brown, M. H., & Sedgewick, R. (1998). Interesting events. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software visualization: Programming as a multi-media experience (pp. 155-172). Cambridge, MA: MIT Press.

CAE SimuFlite. (n.d.) CAE SimuFlite company history. Retrieved February 19, 2003, from http://www.caesimuflite.com/press/history.html

Bruce, L. R., & Sleeman, P.J. (2000). Instructional design: A primer. Greenwich, CT: Information Age Publishing. Bruner, J. (1996). Toward a theory of instruction. New York: Norton. Brusilovsky, P., & Nijhawan, H. (2002, October). A framework for adaptive e-learning based on distributed re-usable learning activities. In M. Driscoll & T. C. Reeves (Eds.), Proceedings of World Conference on E-Learning, E-Learn 2002 (pp. 154-161). Montreal, Canada. Bruvilovsky, P. (2001). Adaptive hypermedia. User Modeling and User-Adapted Interaction, 11, 87-110. Bruvilovsky, P. (2003). Adaptive navigation support in educational hypermedia: The role of student knowledge level and the case for meta-adaptation. British Journal of Educational Technology, 34(4), 487-497. Buch, K., & Bartley, S. (2002). Learning style and training delivery mode preference. Journal of Workplace Learning, 14(1), 5-10.

CAE SimuFlite. (n.d.). CAE company overview. Retrieved January 28, 2003, from http://cae.com/en/about/index. shtml CAE SimuFlite. (n.d.). Power systems and simulation. Retrieved February 19, 2003, from http://www.cae. com/en/power/index.shtml Campbell, D. J. (1988). Task complexity: A review and analysis. Academy of Management Review, 13(1), 40-52. Campbell, J. D., & Mahling, D.E. (1998). A visual language system for developing and presenting Internetbased education. In Proceedings of IEEE Symposium on Visual Languages, (pp. 66-67). Nova Scotia, Canada. Carchiolo, V., Longheu, A., & Malgeri, M. (2002). Adaptive formative paths in a Web-based learning environment. Educational Technology & Society, 5(4). From http://ifets.ieee.org/periodical/vol_4_2002/carchiolo. html Cardus, J. (2006, November). The power of the collective. E.learning Age, pp. 10-12.

Butters, G. (1999, December). Learning At lightspeed. Business 2.0. Retrieved on February 16, 2003, from http:// www.business2.com/articles/mag/0,1640,13310,FF. html

Casella, G., Costagliola, G., Ferrucci, F., Polese, G., & Scanniello, G. (2007). A SCORM thin client architecture for e-learning systems based on Web services. International Journal of Distance Education Technologies, 5(1), 13-30.

CAE SimuFlite. (2002, September 10). CAE SimuFlite Premiers Boeing Business Jet Training Program with CAE Simfinity™ Technology (Press Release). Retrieved February 19, 2003, from http://www.simuflite.com/ press/02_11pr.html

Causen, A., Spanias, A., Xavier, A., & Tampi, M. (1998). A Java signal analysis tool for signal processing experiments. In Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, Vol. 3, (pp. 1849-1852). Seattle, Washington, USA.

0

Compilation of References

Chalmers University of Technology’s I-Lab. (n.d.). Retrieved from http://www.ic.chalmers.se/ilab/ Chang, C. K., Chen, G. D., Liu, B.J., & Ou, K. L. (1996). A language for developing collaborative learning activities on World Wide Web. In Proceedings of 20th International Conference on Computer Software and Applications Conference, Seoul, South Korea (pp. 548-552). Chang, S. K., Arndt, T., Levialdi, S., Liu, A. C., Ma, J., Shih, T., et al. (2000). Macro university: A framework for a federation of virtual universities. International Journal of Computer Processing of Oriental Languages, 13(3), 205-221. Chang, W. F., Wu, Y. C., Chiu, C. W., & Yu, W. C. (2003). Design and implementation of a Web-based distance PLC laboratory. In Proceedings of the 35th Southeastern Symposium on System Theory, (pp. 326-329). Morgantown, West Virginia, USA. Charan, R. (1976). Classroom techniques in teaching by the case method. The Academy of Management Review, 1(3), 116-123. Chiesa, M., & Robertson, A. (2000). Precision teaching and fluency training: Making math easier for pupils and teachers. Educational Psychology in Practice, 16(3), 297-310. Christensen, R., & Hansen, A. (1981). Teaching and the case method. Boston: Harvard Business School Press. Christoffersen, R.D., & Christoffersen, A.H. (1995). Instructional system design: Its role in Coast Guard training and how it can influence the development of training materials. In Proceedings of the IEEE Professional Communication Conference: Smooth Sailing to the Future, Savannah, GA, USA (pp. 39-43). Chu, G. L. (1999). The relationships between cultural differences among American and Chinese university students and the design of personal pages on the World Wide Web. Unpublished doctoral dissertation, University of Georgia.

0

Ciancarini, P. (2002). Coordinating multi-agent applications on the WWW: A reference architecture. IEEE Trans. on Software Engineering, 24(5), 363-375. Compeau, D. R., & Higgins, C. A. (1995). Application of social cognitive theory to training for computer skills. Information Systems Research, 6(2), 118-143. Concepcion, A. I., Ruan, J., & Samson, R. R. (2002). SPIDER: A multi-agent architecture for Internet distributed computing system. In Proceedings of the ISCA 15th International Conference on Parallel and Distributed Computing Systems, (pp. 147-152). Louisville, Kentucky, USA. Cooley, R., Mobasher, B., & Srivastava, J. (1999). Data preparation for mining World Wide Web browsing patterns. Journal of Knowledge and Information Systems, 1(1). Costabile, M., De Marsico, M., Lanzilotti, R., Plantamura, V., & Roselli,T.(2005). On the usability evaluation of e-learning applications. In Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS’05) (Vol. 1, p. 6.2). Washington, DC: IEEE Computer Society. Cox, K. C., & Roman, G. -C. (1992). Abstraction in algorithm animation. In Proceedings of the 1992 IEEE Workshop on Visual Languages, Seattle, WA (pp. 18-24). IEEE Computer Society Press Cristea, A. (2003). Adaptive patterns in authoring of educational adaptive hypermedia. Educational Technology & Society, 6(4). Retrieved July 18, 2006, from http://ifets. ieee.org/periodical/6_4/1.pdf Cristea, A. (2004, February). Adaptive and adaptable educational hypermedia: Where are we now and where are we going? In Proceedings of Web-based Education, Innsbruck, Austria. Retrieved November 8, 2005 from http://wwwis.win.tue.nl/~acristea/HTML/ftp_paper_chrono.html Cronbach, L., & Snow, R. (1977). Aptitudes and instructional methods: A handbook for research on interactions. New York: Irvington.

Compilation of References

Crosby, M.E., & Stelovsky, J. (1995). From multimedia instruction to multimedia evaluation. Journal of Educational Multimedia and Hypermedia, 4(2-3), 147-162. Cuevas, H. M., Fiore, S. M., Bowers, C. A., & Salas, E. (2004). Fostering constructive cognitive and metacognitive activity in computer-based complex task training environments. Computers in Human Behavior, 20(2), 225-241. Cunningham, S., Ryan, Y., Stedman, L., Tapsall, S., Bagdon, K., Flew, T., & Coaldrake, P. (2000). The business of borderless education (pp. 18-23). Canberra: DETYA. Cyber terror the way of the future. (1999, July 13). AAP General News (Australia). Retrieved February 19, 2003, from http://www.elibrary.com/ CyberAtlas. (2003). E-Learning market expanding beyond IT training. Jupiter Research. Retrieve July 28, 2006, from http://cyberatlas.internet.com/markets/education/article/0,,5951_914901,00.html Cynthia, J., Finelli, M., & Wicks A. (2000, May). An instrument for assessing the effectiveness of the circuits curriculum in an electrical engineering program. IEEE Transactions on Education, 43(2), 137-142. Danny, B. L., & Mitsuru, O. (2001). Seven good reasons for mobile agents. Communications of the ACM, 42(3), 86-89. David, K., & Robert, S. G. (2002). Mobile agents and the future of the Internet. ACM Operating Systems Review, 33(3), 7-13. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. Davis, G. A. (1983). Educational psychology. New York: Allyn & Bacon. Decker, P. J., & Nathan, B. R. (1985). Behavior modeling training. New York: Praeger. Deering, S. (1998). Multicast routing in Internetworks and extended LANs. In Proceedings of ACM SIGCOMM (pp. 55-64).

Deitel, H. M., & Deitel, P. J. (1999). Java™ How to program (3rd ed.). Upper Saddle River, NJ: Prentice Hall. Delio, M. (2000, August). Report: Online training ‘boring.’ Wired. Retrieved on February 16, 2003, from http:// www.wired.com/news/print/0,1294,38504,00.html Demmings, B., Shakshuki, E., & Müldner, T. (2007). AETA: Algorithm explanation teaching agent. In Proceedings of the International Conference on Information Technology: New Generations (ITNG), Las Vegas, USA (pp. 616-621). IEEE Computer Society Dennis, W. (2002) Applying what we know: Student learning styles. Retrieved July 18, 2006, from http://www. csrnet.org/csrnet/articles/student-learning-styles.htm Designer’s Edge. (2003). The industry standard instructional design tool. Retrieved from http://www. allencomm.com/products/authoring_design/designer/ Deubel, P. (2003). An investigation of behaviorist and cognitive approaches to instructional multimedia design. Journal of Educational Multimedia and Hypermedia, 12(1), 63-90. Deusdado, S. (2002). Integração Adaptativa de Aplicações Multicast para Conferência Multimédia. Master Thesis, Departamento de Informática, Universidade do Minho. Dey, A. K. (2000). Enabling the use of context in interactive applications. In Proceedings of the 2000 Conference on Human Factors in Computing Systems, (pp. 79-80). Hague, The Netherlands. Diehl, S. (Ed.). (2002). Software visualization (LNCS State-of-the-Art Survey 2269). Springer. Diehl, S., & Kerren, A. (2002). Reification of program points for visual execution. In Proceedings of the First IEEE International Workshop on Visualizing Software for Understanding and Analysis (VisSoft ’02) (pp. 100109). IEEE Computing Society Press. Diehl, S., Görg, C., & Kerren, A. (2002). Animating algorithms live and post mortem. In S. Diehl (Ed.), Software visualization (LNCS State-of-the-Art Survey 2269, pp. 46-57). Springer.

0

Compilation of References

Diestel, R. (2000).Graph theory (2nd ed.). Germany: Springer.

Proceedings of the ASEE/IEEE Frontiers in Education Conference, Reno, NV, USA (Vol. 3, pp. F4E-1-5).

Dijkstra, E.W. (1989). On the cruelty of really teaching computing science. The SIGCSE Award Lecture. Communications of the ACM, 31(12), 1398-1404.

Dow, C. R, Lin, C. Y., & Hsu, F. W. (2002). A mobile agent-based virtual language learning laboratory. In Proceedings of the International Conference on Chinese Language Computing, (pp. 98-103). Taichung, Taiwan.

Dillon, H.C. (2007). Military training for civilian careers (Or: How to gain practical experience while serving your country). Occupational Outlook Quarterly, 51(1), 7-17. Dirchinck-Holmfeld, L., & Lorentsen, A. (2003). Transforming university practice through ICT – Integrated perspectives on organizational, technological, and pedagogical change. Interactive Learning Environments, 11(2), 91-110. Dix, A., Finlay, F., Abowd, G., & Beale, R. (1998). Human computer interaction (2nd ed.). England: Prentice Hall Europe. DMATEK Co. Ltd. (n.d.). Retrieved from http://www. dmatek.com.tw Domingo, P., Garcia-Crespo, A., Ruiz, B., & Iglesias, A. (2002). Edu-Ex: A tool for auto-regulated intelligent tutoring systems development based on models. Avenida de la Universidad, 30.28911, Artificial Intelligence Review 18. Kluwer Academic. Donati, J. (2003). Least-squares deconvolution of Stellar Spectra. Retrieved July 28, 2006, from http://webast.ast. obs-mip.fr/people/donati/multi.html Dorca, F. A., Lopes, C. R., & Fernandes, M. A. (2003). A multiagent architecture for distance education systems. In Proceedings of the 3rd IEEE International Conference on Advanced Learning Technologies (pp. 368-369), Athens, Greece. Dospisil, J., & Kendall, E. (2001). Design agents with negotiation capabilities. In S.M. Rahman & R.J. Bignall (Eds.), Internet commerce and software agents (pp. 122149). Hershey, PA: Idea Group Publishing. Douglas, I. (2001). Instructional design based on reusable learning object: Applying lessons of object-oriented software engineering to learning system design. In

0

Dow, C. R., Lin, C. Y., Shen, C. C., Lin, J. H., & Chen, S. C. (2002). A virtual laboratory for macro universities using mobile agent techniques. The International Journal of Computer Processing of Oriental Languages, 15(1), 1-18. Downes, S., Babin, G., Belliveau, L., Blanchard, R., Levy, G., Bernard, P., Paquette, G., & Plourde, S. (2003, April 22–23). Distributed digital rights management: The EduSource approach to DRM. In Proceedings of the First International ODRL Workshop, Vienna. Dunn, R. (1996). How to implement and supervise a learning style program. VA: ACSD. Dunn, R., Dunn K., & Price, G. (1989). Productivity environmental preference survey. Lawrence, KS: Price Systems. Dunn, R., Griggs, S. A., Olson, J., Gorman, B., & Beasley, M. (1995). A meta-analytic validation of the Dunn and Dunn learning styles model. Journal of Educational Research, 88(6), 353-361. Durham, A. G., & Emurian, H. H. (1998). Learning and retention with a menu and a command line interface. Computers in Human Behavior, 14, 597-620. Eades, P., & Zhang, K. (Eds.). (1996). Software visualization. Singapore: World Scientific Pub. Eilam, B., & Aharon, I. (2003). Student’s planning in the process of self-regulated learning. Contemporary Educational Psychology, 28(3), 304-334. Eklund, J., Kay, M., & Lynch, H. (2003). E-learning: Emerging issues and key trends. Australian Flexible Learning Framework, Australian National Training Authority.

Compilation of References

elearningpost. (n.d.). Retrieved from http://www.elearningpost.com Elmasri, R., & Navathe, S.B. (2007). Fundamentals of database systems (5th Ed.). Addison-Wesley. Emurian, H. H. (2004). A programmed instruction tutoring system for Java™: Consideration of learning performance and software self-efficacy. Computers in Human Behavior, 20(3), 423-459. Emurian, H. H. (2005). Web-based programmed instruction: Evidence of rule-governed learning. Computers in Human Behavior, 21, 893-915. Emurian, H. H., & Durham, A. G. (2001, May 20-23). A personalized system of instruction for teaching Java. In M. Khosrow-Pour (Ed.), Managing information Technology in a Global Economy, 2001 Information Resources Management Association International Conference (pp. 155-160), Toronto, Ontario, Canada. Hershey, PA: Idea Group Publishing. Emurian, H. H., & Durham, A. G. (2002, May 19-22). Enhanced learning on a programmed instruction tutoring system for JAVA. In M. Khosrow-Pour (Ed.), Issues & Trends of Information Technology Management in Contemporary Organizations, 2002 Information Resources Management Association International Conference (pp. 205-208), Seattle, Washington, USA. Hershey, PA: Idea Group Publishing. Emurian, H. H., & Durham, A. G. (2003). Computerbased tutoring systems: A behavioral approach. In J. A. Jacko, & A. Sears (Eds.), Handbook of human-computer interaction (pp. 677-697). Mahwah, NJ: Lawrence Erlbaum & Associates. Emurian, H. H., Hu, X., Wang, J., & Durham, A. G. (2000). Learning Java: A programmed instruction approach using applets. Computers in Human Behavior, 16, 395-422. Emurian, H. H., Wang, J., & Durham, A. G. (2003). Analysis of learner performance on a tutoring system for Java. In T. McGill (Ed.), Current issues in IT education (pp. 46-76). Hershey, PA: IRM Press.

Engeström, Y. (1987). Learning by expanding: An activity-theoretical approach to developmental research. Helsinki: Orienta-Konsultit. eXist. (2007). Open source native XML database. Retrieved from http://exist.sourceforge.net/ Faltin, N. (2001). Structure and constraints in interactive exploratory algorithm learning. In S. Diehl (Ed.), Software visualization - international seminar (Rev. Papers) (LNCS 2269, pp. 213-226). Dagstuhl Castle, Germany.. Feng, Z. (2003). Algorithm visualization: Using various levels of abstraction and invariants. Honours Thesis, Acadia University. Wolfville, Nova Scotia, Canada Fensel, D. (2001). Ontologies: Silver bullet for knowledge management and electronic commerce. Berlin: Springer Verlag. Fensel, D., Motta, E., Decker, S., & Zdrahal, Z. (1997). Using ontologies for defining tasks, problem-solving methods and their mappings. In E. Plaza et al. (Eds.), Knowledge acquisition, modeling and management. Springer-Verlag. Lecture Notes in Artificial Intelligence (LNAI) 1319. Fernandez, G., Zhao, L., & Wijiegunaratne, I. (2003). Design patterns for a federation architecture. Journal of Object Technology, 2(3), 135-149. Ferrucci, F., Tortora, G., & Vitiello, G. (2002). Visual programming. In Encyclopaedia of Software Engineering. New York: John Wiley & Sons. Fetscherin, M., & Schmid, M. (2003). Comparing the usage of digital rights management systems in the music, film, and print industry. In Proceedings of the ICEC 2003. Pittsburgh, Pennsylvania. Fleischer, R., & Kucera, L. (2001). Algorithm animation for teaching. In S. Diehl (Ed.), Software Visualization - International Seminar (Rev. Papers) (LNCS 2269, pp. 113-128). Dagstuhl Castle, Germany. Fok, A., & Ip, H. (2004). Personalized Education (PE) — An exploratory study of learning pedagogies in relation to personalization technologies. In Proceedings



Compilation of References

of Advances in Web-Based Learning (ICWL 2004). Lecture Notes in Computer Science (LNCS 3143) (pp. 407-415).

Gamma, E., Helm, R., Johnson, J., & Vlissides, J. (1995). Design patterns. Elements of reusable object-oriented software. Reading, MA: Addison-Wesley.

Ford, N. (2000). Cognitive styles and virtual environments. Journal of the American Society for Information Science, 51(6), 543-557.

Gan, W. S., Chong, Y. K., Gong, W., & Tan, W. T. (2000). Rapid prototyping system for teaching real-time digital signal processing. IEEE Transactions on Education, 43(1), 19-24.

Fowler, M. (2000). UML distilled (2nd Ed.). Reading, MA: Addison-Wesley. Fox, M. A., & Hackerman, N. (2003). Evaluating and improving undergraduate teaching in science, technology, engineering, and mathematics. Washington, DC: The National Academies Press.

Ganimal. (2007). Project homepage. Retrieved from http://www.cs.uni-sb.de/GANIMAL Garcia-Molina, H., Ullman, J., & Widom, J. (2000). Database system implementation. New Jersey: Prentice Hall.

Franklin, D., Flachsbart, J., & Hammond, K., (1999). The intelligent classroom. IEEE Intelligent Systems Journal, 14(5), 2-5.

Garlan, D., & Siewiorek, D. P. (2002). Project aura: Toward distraction-free pervasive computing. IEEE Pervasive Computing, 1(2), 22-31.

Fridman, N. N., & McGuinness, D. L. (2001, March). Ontology development 101: A guide to creating your first ontology (Stanford Knowledge Systems Laboratory Technical Report KSL-01-05, Stanford Medical Informatics Technical Report SMI-2001-0880).

Gemini Learning Systems Inc. (n.d.). Retrieved from http://www.gemini.com/

Friedman, J., Bentley, J., & Finkel, R. (1977). An algorithm for finding best matches in logarithmic expected time. ACM Transactions on Mathematical Software, 3(3), 209-226. Friedman, T. L. (2005). The world is flat: A brief history of the twenty-first century. New York: Farrar, Straus, & Giroux. Fuji-shi. (2005). Fuji-shi live camera. Retrieved from http://www.city.fuji.shizuoka.jp/live/ Fulk, J. (1993). Social construction of communication technology. Academy of Management Journal, 36, 921-950. Gagne, M., & Deci, E. (2005). Self-determination theory and work motivation. Journal of Organizational Behaviour, 26, 331-362. Gagne, R. M. (1992). Principles of instructional design. New York: Holt, Rinehart and Winston, Inc.



Gilliland-Swetland, A. J. (2000). Introduction to metadata, setting the stage. Retrieved December 20, 2004, from http://www.getty.edu/research/institute/standards/ intrometadata/ Gist, M. E., Schwoerer, C., & Rosen, B. (1989). Effects of alternative training methods on self-efficacy and performance in computer software training. Journal of Applied Psychology, 74, 884-891. Gloor, P.A. (1992). AACE - algorithm animation for computer science education. In Proceedings of the 1992 IEEE Workshop on Visual Languages, Seattle, WA (pp. 25-31). Gloor, P.A. (1998). Animated algorithms. In J. Stasko, J. Domingue, M.H. Brown, B.A. Price (Eds.), Software visualization: Programming as a multimedia experience (pp. 409-416). Cambridge, MA: MIT Press. Gloor, P.A. (1998). User interface issues for algorithm animation. In J. Stasko, J. Domingue, M.H. Brown, B.A. Price (Eds.), Software visualization: Programming as a multimedia experience (pp. 145-152). Cambridge, MA: MIT Press.

Compilation of References

González-Castaño, F. J., Anido-Rifón, L., Vales-Alonso, J., Fernández-Iglesias, M. J., Llamas Nistal, M., Rodríguez-Hernández, P., & Pousada-Carballo, J. M. (2001). Internet access to real equipment at computer architecture laboratories using the ava/CORBA paradigm. Computers & Education, 36(2), 151-170. Goodrich, M., & Tamassia, R. (2001). Data structures and algorithms in Java (2nd ed.). New York: John Wiley & Sons, Inc. Goodyear, P. (1997). Instructional design environments: Methods and tools for the design of complex instructional systems. In S. Dijkstra, N. Seel, F. Schott & R. Tennyson (Eds.), Instructional design: International perspectives (pp. 83-111). Mahwah NJ: Lawrence Erlbaum Associates. Goodyear, P. (1999). Seeing learning as work: Implications for understanding and improving analysis and design. Journal of Courseware Engineering, 2, 3-11. Gosling, J., & Mintzberg, H. (2006). Management education as if both matter. Management Learning, 37(4), 419-428 Grau, B. C. (2004). A possible simplification of the semantic Web architecture. In Proceedings of the WWW2004. New York. Greenburg, P. (2001). CRM at the speed of light. McGraw Hill. Greer, R. D. (2002). Designing teaching strategies: An applied behavior analysis systems approach. New York: Academic Press. Greer, R. D., & McDonough, S. H. (1999). Is the learn unit a fundamental measure of pedagogy? The Behavior Analyst, 22, 5-16. Gruber, T. R. (1993). A translation approach to portable ontologies. Knowledge Acquisition, 5(2), 199-220. Gruber, T. R. (1993, March). Toward principles for the design of ontologies used for knowledge sharing. In Padua Workshop on Formal Ontology (p. 23).

Guarino, N. (1998). Formal ontology and information systems. In Proceedings of FOIS ‘98, Trento, Italy. Guimaraes, E., Maffeis, A., Pereira, J., Russo, B., Cardozo, E., Bergerman, M., et al. (2003). REAL: A virtual laboratory for mobile robot experiments. IEEE Transactions on Education, 46(1), 37-42. Gururajan, V. (2002). Knowledge management a new game in higher education: A discussion paper on the new trends of technology influencing the higher education landscape. In Focusing on the Student. Proceedings of the 11th Annual Teaching Learning Forum. Perth: Edith Cowan University. Retrieved January 2007, from http://lsn.curtin.edu.au/tlf/tlf2002/abstracts/gururajanabs.html Guth, S. (2003). Rights expression languages. In E. Becker, W. Buhse, D. Günnewig, & N. Rump (Eds.), Digital rights management: Technological, economic, legal and political aspects (pp. 101-112). Heidelberg, Germany: Springer. Guth, S., Neumann,G., & Strembeck, M.(2003). Experiences with the enforcement of access rights extracted from ODRL-based digital contracts. In Proceedings of the ACM Workshop on Digital Rights Management, Washington, DC. Hackman, J. R. (1968). Effects of task characteristics on group products. Journal of Experimental Social Psychology, 4(2), 162-187. Hansen, S.R., Narayanan, N.H., & Hegarty, M. (2002). Designing educationally effective algorithm visualizations: Embedding analogies and animations in hypermedia. Journal of Visual Languages and Computing, 13(3), 291-317. Hardman, V., Kirstein, P., Sasse, A., Handley, M., & Watson, A. (1995). RAT, Robust Audio Tool. Retrieved from http://www-mice.cs.ucl.ac.uk/multimedia/software/rat/ Harley, H. D., Seals, C. D., & Rosson, M. B. (1998, Winter). A formative evaluation of scenario-based tools for learning object-oriented design. ACM Crossroads,



Compilation of References

1-5. Retrieved April 6, 2004, from http://www.acm. org/crossroads.xrds5-1/eval.html

Object Metadata (LOM). In Proceedings of ED-MEDIA 2001, Tampere, Finland.

Hayes, P. (2003). RDF semantics. World Wide Web Consortium. Retrieved July 27, 2006, from http://www. w3.org/TR/2003/PR-rdf-mt-20031215/

Honey, P., & Mumford, A. (1986). Using your learning styles. Maidenhead, Berkshire: Honey Publications.

Hayes, S. C. (1989). Rule-governed behavior: Cognition, contingencies, and instructional control. New York: Plenum Press. Hayes, S. C., Fox, E., Gifford, E. V., Wilson, K. G., Barnes-Holmes, D., & Healy, O. (2001). Derived relational responding as learned behavior. In S. C. Hayes, D. Barnes-Holmes & B. Roche (Eds.), Relational frame theory: A post-skinnerian account of human language and cognition. New York: Kluwer Academic. Herold, D.M., Davis, W., Fedor, D.B., & Parsons, C.K. (2002). Dispositional influences on transfer of learning in multistate training programs. Personnel Psychology, 55(4), 851-869. Hesselink, L., Rizal, D., Bjornson, E., Paik, S., Batra, R., Catrysse, P., et al. (2003). Standford cyberLab: Internet assisted laboratories. Journal of Distance Education Technologies, 1(1), 22-39. Hitachi Hybrid Network. (2004). IP visual communication system NetCS. Retrieved from http://www.hitachihybrid.co.jp/business/net/netcs01.html Hoadley, C. M. (2004). Methodological alignment in design-based research. Educational Psychologist, 39(4), 203-212. Hoffman, C., & Goodwin, S. (2006). A clicker for your thoughts: Technology for active learning. New Library World, 107(9/10), 422 Hofstede, G. (1997) Cultures and organizations: Software of the mind. New York: McGraw-Hill. Holbrook, H., & Cain, B. (2006). Source-specific multicast for IP (RFC 4607, IETF). Network Working Group. Holzinger, A., Kleinberger, T., & Müller, P. (2001). Multimedia learning systems based on IEEE Learning



Hong, H., & Kinshuk. (2004, June). Adaptation to student learning styles in web-based educational systems. In L. Cantoni & C. Mcloughlin (Eds.), Proceedings of ED-MEDIA 2004 - World Conference on Educational Multimedia, Hypermedia & Telecommunications (pp. 491-496), Lugano, Switzerland. Horstmann, C. (2001). Big Java: Programming and practice. New York: John Wiley & Sons. Horton, W. (2000). Designing Web-based training: How to teach anyone anything anywhere anytime. New York: John Wiley & Sons. Howell, S.L., Williams, P.B., & Lindsay, N.K. (2004). Thirty-two trends affecting distance education: An informed foundation for strategic planning. Online Journal of Distance Learning Administration, 6(3). Retrieved March 2006, from http://www.emich.edu/cfid/ PDFs/32Trends.pdf Hsu, S., & White, J. (1998). Developing holistic technology-based training programs. Retrieved February 17, 2003, from www.viats.org/papers/session5a2.htm, 1, 2 Huang, Z., & Cappel, J. R. (2005). Assessment of a Web-based learning game in an information systems course. The Journal of Computer Information Systems, 45(4), 42-49 Hundhausen, C.D., Douglas, S.A., & Stasko, J.T. (2002). A meta-study of algorithm visualization effectiveness. Journal of Visual Languages and Computing, 13(3), 259-290. Hunter, D. (2008). The virtual student/client experience. Journal of American Academy of Business, 12(1), 8892. Iannella, R. (2001). Digital Rights Management (DRM) architectures. D-Lib Magazine, 7(6).

Compilation of References

IBM’s Java Aglet. (n.d.). Retrieved from http://www.trl. ibm.com/aglets IDC. (2002, September 30). While corporate training markets will not live up to earlier forecasts, IDC suggests reasons for optimism, particularly e-learning. Retrieved July 28, 2006, from http://www.idc.com/getdoc. jhtml?containerId=pr2002_ 09_17_150550

Ives, B., Olson, M., & Baroudi, S. (1983). The measurement of user information satisfaction. Communications of the ACM, 26, 785-793. Jannink, J. (1999, July). Ontology maintenance with an algebraic methodology: A case study. In Proceedings of 1999 AAAI Workshop on Ontology Management, Orlando, Florida.

IEEE Learning Technology Standards Committee (LTSC). (n.d.). Retrieved from http://ltsc.ieee.org

Jarc, D.J. (1995). Interactive data structure visualizations. Retrieved from http://nova.umuc.edu/~jarc/idsv

Ikeda, M., Seta, K., Kakusho, O., & Mizoguchi, R. (1998). Task ontology: Ontology for building conceptual problem solving models. In Proceedings of the 13th European Conference on Artificial Intelligence ECAI’98.

JMF 2.0 (1999). JMF - Java TM Media Framework API Guide. Sun Microsystems, JMF 2.0 FCS.

IMG Global Learning Consortium (IMG). (n.d.). Retrieved from http://www.imsglobal.org IMS. (2003). IMS learning design best practice and implementation guide, version 1.0 final specification. Inoue, T., Okada, K., & Matsushita, Y. (1997). Spatial design for integration of face-to-face and video meetings: HERMES video conferencing system. IEICE, J80-DII(9), 2482-2492. (in Japanese) Instone, K. (2000). Information architecture and personalization. Argus Associates. Retrieved from http://argusacia.com/white_papers/personalization.html International Data Corporation. (2002, September 30). While corporate training markets will not live up to earlier forecasts, IDC suggests reasons for optimism, particularly eLearning. International Data Corporation. Retrieved March 5, 2003, from http://www.idc.com/getdoc.jhtml?containerId=pr2002_09_17_150550 ITU-JPEG. (1992). JPEG Standard, Information Technology - digital compression and coding of continuous-tone still images - requirements and guidelines. Recommendation T.81, ITU. ITU-T H.261. (1993). ITU-T Recommendation H.261: Video CODEC for audiovisual services At p x 64 kbits. ITU-T H.263. (1998). ITU-T Recommendation H.263: Video coding for low bitrate communication.

Johannes, D., Seidel, R., & Wilhelm, R. (2005). Algorithm animation using shape analysis: Visualizing abstract executions. In Proceedings of the ACM Symposium on Software Visualization (SoftVis ’05), St. Louis, MO, USA (pp. 17-26). ACM. Johanson, M., & Lie, A. (2002). Layered encoding and transmission of video in heterogeneous environments. ACM. Johnson, R. A., & Wichern, D. W. (1992). Applied multivariate statistical analysis (3rd ed.). New Jersey: Prentice Hall. Jokela, S. (2001). Metadata enhanced content management in media companies. In Acta Polytecnica Scandinavica. Mathematics and computing series no. 114. Doctoral thesis, Helsinki University of Technology. Jonassen, D. H., & Grabowski, B. (1993). Individual differences and instruction. Computers in Human Behaviour, 11(3-4), 371-390. Jones, C., Connolly, M., Gear, A., & Read, M. (2006). Collaborative learning with group interactive technology: A case study with postgraduate students. Management Learning, 37(3), 377-396 Jones, D., Shirley, G., & Lynch, T. (2003). An information systems design theory for Web-based education. In Proceedings of the IASTED International Symposium on Web-Based Education, Greece.



Compilation of References

Jsamba. (2007). Project homepage. Retrieved from http://www-static.cc.gatech.edu/gvu/softviz/algoanim/ jsamba/ Jung, J.L., Jacques, P.A., de Andrade, A.F., & Vicari, R.M. (2002). The conception of agents as part of a social model of distance learning. In G. Bittencourt & G.L. Ramalho (Eds.), Advances in artificial intelligence (pp. 140-151). Berlin: Springer-Verlag. Karnik, N. M. (1991). Design Issues in mobile agent programming systems. IEEE Concurrency, 6(3), 125-132. Karp, M., Condit, D., & Nullmeyer, R. (1999). F-16 Cockpit/Crew Resource Management. NBAA Business Aviation Fact Book (2002) (AL/HR-TR-1999-XXXX, in progress). Mesa, AZ: Air Force Research Laboratory.. Retrieved January 28, 2003, from www.nbaa.org/factbook/2002/section4.htm#02 Karp, M.R., McCurry, W.K., Turney, M.A., & Harms, D. (2000). Aviation education for future airline pilots: An integrated model. Retrieved February 17, 2003, from www.viats.org/papers/session5d4.htm, 1, 3

Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral research. New York: Harcourt. Kerren, A. (2004). Generation as method for explorative learning in computer science education. In Proceedings of the 9th Annual Conference on Innovation and Technology in Computer Science Education (ITiCSE ’04), (pp. 77-81). Leeds, UK: ACM Press. Kerren, A. (2004). Learning by generation in computer science education. Journal of Computer Science and Technology (JCS&T), 4(2), 84-90. Kerren, A., & Stasko, J.T. (2002). Algorithm animation – chapter introduction. In S. Diehl (Ed.), Software visualization (LNCS State-of-the-Art Survey 2269, pp. 1-15). Springer. Kerren, A., Müldner, T., & Shakshuki, E. (2006). Novel algorithm explanation techniques for improving algorithm teaching. In Proceedings of the 3rd ACM Symposium on Software Visualization (SoftVis ’06), (pp. 175-176). Brighton, UK. ACM Press

Kasowitz, A. (1997). Tool for automating instructional design. ERIC Educational Reports. Retrieved from http://ericit.org/digests/EDO-IR-1998-01.shtml

Khalifa, M., & Liu, V. (2008). Semantic network representation of computer-mediated discussions: Conceptual facilitation form and knowledge acquisition. Omega, 36(2), 252.

Kassanke, S., El-Saddik, A., & Steinacker, A. (2001). Learning objects metadata and tools in the area of operations research. In Proceedings of ED-MEDIA 2001, Tampere, Finland.

Khopkar, Y., Spink, A., Giles, C. L., Shah, P., & Debnath, S. (2003). Search engine personalization. Retrieved September 30, 2005, from http://firstmonday.org/issues/issue8_7/khopkar/index.html

Kato, V., Jiang, D., & Hakozaki, K. (2003). A proposal of a streaming video system in best-effort networks using adaptive QoS control rules. In Proceedings of IPSJ DPSWS’2003. IPSJ Symposium Series, 2003(19) (pp. 7-12). (in Japanese)

Kimovski, G., Trajkovic, V., & Davcev, D. (2003). Negotiation-based multiagent resource management in distance education. In V. Uskov (Ed.), Computers and advanced technology in education (CATE) (pp. 36-41). Anaheim, CA: ACTA Press.

Kawamura, N., Maita, Y., Hashimoto, K., & Shibata, Y. (2003). Remote medical support system using the transcoding function (IPSJ SIG Technical Report DPS 113-5), pp. 69-74. (in Japanese)

Kinshuk, Patel, A., Oppermann, R., & Russell, D. (2001). Role of human teacher in Web-based intelligent tutoring systems. Journal of Distance Learning, 6 (1), 26-35.

Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral research. New York: Harcourt Brace College Publishers.



Kirpatrick, D. L. (Ed.). (1967). Evaluation of training: Training and development handbook. New York: McGraw-Hill.

Compilation of References

Ko, C. C., Chen, B. M., Hu, S., Ramakrishnan, V., Cheng, C. D., Zhuang, Y., et al. (2001). A Web-based virtual laboratory on a frequency modulation experiment. IEEE Transactions on Systems, Man and Cybernetics, 31(3), 295-303. Kolb, D. (1984). Experiential learning. Englewood Cliffs, NJ: Prentice Hall. Kosiur, D. (1998). IP multicasting: The complete guide to interactive corporate networks. New York: John Wiley & Sons, Inc. Kotsiopulos, A., & Hartley, N. (2004). Leveraging multi-institutional resources: GP-IDEA – A collaborative distance education model. In Proceedings of the 21st ICDE World Conference on Open Learning & Distance Education, Lifelong Learning in the Networked World, Hong Kong. Kudadjie-Gyamfi, E., & Rachlin, H. (2002). Rulegoverned versus contingency-governed behavior in a self-control task: Effects of changes in contingencies. Behavioral Processes, 57(1), 29-35. Kumamoto-shi (2001). Today’s Kumamoto Castle. Retrieved from http://www.city.kumamoto.kumamoto.jp/ castleToday.html Kurki, T., Jokela, S., Sulonen, R., & Turpeinen, M., (1998). Agents in delivering personalized content based on semantic metadata. American Association for Artificial Intelligence. Retrieved September 30, 2005, from http://www.aaai.org/ Laister, J., & Koubek, A. (2001). 3rd generation learning platforms requirements and motivation for collaborative learning. European Journal of Open and Distance Learning. Retrieved from http://www.eurodl.org/materials/contrib/2001/icl01/laister.htm Lamminaho, V. (2000). Metadata specification: Forms, menus for description of courses and all other objects. CUBER project, Deliverable D3.1. Lange, D. B., & Oshima, M. (1998). Programming and deploying Java mobile agents with aglets. Boston: Addison Wesley.

Laudon, K. C., & Laudon, J.P. (2002). Management information systems. Prentice-Hall. Lavooy, M. J., & Newlin, M. H. (2003). Computer mediated communication: Online instruction and interactivity. Journal of Interactive Learning Research, 14(2), 157-165. Lawn, M. (2001). Borderless education: Imagining a European education space in a time of brands and networks. Disprogram, 22(2), 173-184. Layaida, O., & Hagimonte, D. (2002). Dynamic adaptation in distributed multimedia applications (Technical Report). INRIA. Lederer, A. L., Maupin, D. J., Maupin, D. J., Sena, M.P., & Zhuang, Y. (2000). The technology acceptance model and the World Wide Web. Decision Support Systems, (29), 269-282. Leher, H.R., Moore, P.J., & Telfer, R.A. (1999). Approaches to learning: Is there a better way to prepare future pilots? Retrieved February 17, 2003, from www. viats.org/papers/session5d1.htm 1-2, 5 Leidner, D. E., & Jarvenpaa, S. L. (1995). The use of information technology to enhance management school education: A theoretical view. MIS Quarterly, 19, 265-291. Levene, H. (1960). In I. Olkin et al. (Eds.) Contributions to probability and statistics: Essays in honor of Harold Hotelling. (pp. 278-292). Stanford University Press. Lewin, K. (1951). Field theory in social science: Selected theoretical papers. New York: Harper and Row. Li, B., Xu, D., Naharstedt, K., & Liu, J. (1998). Endto-end QoS support for adaptive applications over the Internet. Department of Computer Science, University of Illinois at Urbana-Champaign. Lin, J., Ho, C., Sadiq, W., & Orlowska, M.E. (2002). Using workflow technology to manage flexible e-learning services. In Educational Technology & Society, 5(4). Retrieved from http://ifets.ieee.org/periodical/ vol_4_2002/lin.html



Compilation of References

Liu, J., Chan, S., Hung, A., & Lee, R. (2002). Facilitators and inhibitors of e-learning. In L. C. Jain, R. J. Howlett, N. S. Ichalkaranje, & G. Tonfoni (Eds.), Virtual environments for teaching and learning, Series on innovative intelligence (Vol. 1, pp. 75-109). World Scientific. Liu, J., Li, B., & Zhang, Y. (2004). An end-to-end adaptation protocol for layered video multicast using optimal rate allocation. IEEE Transactions On Multimedia, 6(1). Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation. American Psychologist, 57(9), 705-717. Locke, E. A., Chah, D., Harrison, S., & Lustgarten, N. (1989). Separating the effects of goal specificity from goal level. Organizational Behavior and Human Performance, 43, 270-287. Lu, J. L., & Chen, R. F. (2003). Design and implementation of a fine-grained menu control processor for Web-based information systems. Future Generation Computer Systems, pp. 1105-1119. Lubonski, M., Gay, V., & Simmonds, A. (2005). A conceptual architecture for adaptation in remote desktop systems driven by the user perception of multimedia. In Proceedings of the 2005 Asia-Pacific Conference on Communications (pp. 891-895). Macromedia Flash (2007). Macromedia Flash MX. Retrieved from http://www.macromedia.com/software/flash/ Macromedia Inc. (n.d.). Retrieved from http://www. macromedia.com Macromedia. (2003). Flash Communication Server MX 1.5. Retrieved from http://www.macromedia.com/software/ flashcom/ Macromedia. (2004). Flash MX 2004. Retrieved from http://www.macromedia.com/software/flash/ Maeda, K., Aibara, R., Kawamoto, K., Terauchi, M., Kohno, E., & Nishimura, K. (1997). Multimedia communication environment for distance learning. IEICE, J80-B-I(6), 348-354. (in Japanese)



Magoulas, G., Papanikolaou, K., & Grigoriadou, M. (2003). Adaptive Web-based learning: Accommodating individual differences through system’s adaptation. British Journal of Educational Technology, 34(4). Retrieved July 19, 2006, from http://www.dcs.bbk. ac.uk/~gmagoulas/bjet.pdf Marginson, S. (2002). The phenomenal rise of international degrees Down Under. Change, 34(3), 34-43. Retrieved January 2007, from http://www.findarticles. com/p/articles/mi_m1254/is_3_34/ai_85465152 Massachusetts Institute of Technology’s iLab project. (n.d.). Retrieved from http://ilabserv.mit.edu/ilab/ Mauro, R., & Barshi, I. (n.d.). Using an Internet-based decision research system in aviation training research. Retrieved February 17, 2003 from www.viats.org/papers/session5c2.htm Maxwell, S. E., & Delaney, H. D. (2000). Designing experiments and analyzing data. Mahwah, NJ: Lawrence Erlbaum Associates. Mayer, R. E. (2002). The promise of educational psychology. Volume II. Teaching for meaningful learning. Upper Saddle River, NJ: Pearson Education. Mayer, R. E. (2004). Should there be a three-strikes rule against pure discovery learning? The case for guided methods of instruction. American Psychologist, 59(1), 14-19. Mayer, R. E. (Ed.). (1988). Teaching and learning computer programming. Hillsdale, NJ: Erlbaum. McCanne S., & Jacobson, V. (1995). vic: A flexible framework for packet video. In Proceedings of the ACM Multimedia Conference (pp. 511-522). Retrieved from http://www-mice.cs.ucl.ac.uk/multimedia/software/vic/ McCray, R. A., DeHaan, R. L., & Schuck, J. A. (Eds.). (2003). Improving undergraduate instruction in science, technology, engineering, and mathematics. Washington, DC: The National Academies Press. McGehee, W., & Tullar, W. (1978). A note on evaluating behavior modification and behavior modeling as

Compilation of References

industrial training techniques. Personal Psychology, 31, 477-484. Mcloughlin, C. (1999). The implications of research literature on learning styles for the design of instructional material. Australian Journal of Educational Technology, 15(3), 222-241. Retrieved July 19, 2006, from http://www. ascilite.org.au/ajet/ajet15/mcloughlin.htm Mennecke, B. E., Crossland, M. D., & Killingsworth, B. L. (2000). Is a map more than a picture? The role of SDSS technology, subject characteristics, and problem complexity on map reading and problem solving. MIS Quarterly, 24(4), 601-627. Micarelli, A., & Sciarrone, F. (2004). Anatomy and empirical evaluation of an adaptive Web-based information filtering system. User Modeling and User-Adapted Interaction, 14, 159-200. Microsoft. (2004). Netmeeting. Retrieved from http:// www.microsoft.com/windows/netmeeting/ Middlehurst, R. (2003). The developing world of borderless higher education: Markets, providers, quality assurance and qualifications. In Conference Proceedings of the First Global Forum on International Quality Assurance, Accreditation and the Recognition of Qualifications (pp. 25-39). Paris, UNESCO. Miliszewska, I. (2007). Is it fully ‘on’ or partly ‘off’? The case of fully-online provision of transnational education. Journal of Information Technology Education, 6, 499-514. Milojicic, D. (2002). Mobile agent applications. IEEE Concurrency, 7(3), 80-90.

Mobasher, B., Cooley, R., & Srivastava, J. (2000). Automatic personalization based on Web usage mining. Communications of the ACM, 43(8), 142-151. Mohsin, M., Wong, W., & Bhutt, Y. (2001). Support for real-time traffic in the Internet, and QoS issues. Department of Computer Science, University of Texas at Dallas. Moon, J., & Kim, Y. (2001). Extending the TAM for a World Wide Web context. Information & Management, 38, 217-230. Moore, G., & Benbasat, I. (1991). Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research, 2(3), 192-222. Mori, K., Oyabu, Y., Nomura, A., & Oshita, S. (1992). Inter-campus tele-education and its evaluation. IEICE, J75-A(2), 244-255. (in Japanese) Morris, D., Shaw, B., & Perney, J. (1990). Helping low readers in grades 2 and 3: An after-school volunteer tutoring program. The Elementary School Journal, 91, 133-150. Mousavinezhad, S. H., & Abdel-Qader, I. M. (2001). Digital signal processing in theory and practice. In Proceedings of the 31st ASEE/IEEE Frontiers in Education Conference (pp. T2C-13-T2C-16). Reno, Nevada, USA. Müldner, T., & Shakshuki, E. (2006). Explaining algorithms: A new perspective. International Journal of Distance Education Technologies (JDET), 4(3), 6-23.

Miras, D. (2002). A survey of network QoS needs of advanced Internet applications (Working Document). Computer Science Department, University College London.

Müldner, T., Shakshuki, E., & Merill, J. (2004). Selecting media for explaining algorithms. In Proceedings of the AACE World Conference on Educational Multimedia, Hypermedia and Telecommunications (EDMEDIA ’04) (pp. 2048-2053). Lugano, Swizerland: AACE.

Miyoshi, K., Okanaga, Y., Kou, S., & Kondo, S. (2000). Design, development, and experiments of distance learning systems by satellite Internet services. IEICE, J83-D-I(6), 644-650. (in Japanese)

Muraida, D.J., & Spector, J.M. (1997). Automatic design instruction. In S. Dijkstra, N. Seel, F. Schott, & D. Tennyson (Eds.), Instructional design: International perspectives (Vol. 2). Mahwah, NJ: Lawrence Erlbaum.



Compilation of References

Musser D.R., & Saini, A. (1996). STL tutorial and reference guide: C++ programming with the standard template library. Reading, MA: Addison-Wesley.

NTT West. (2004). Technical Report 12. Retrieved from http://www.ntt-west.co.jp/ipnet/ip/siryou.pdf (in Japanese)

Nacimiento Inc. (2001). Virtual instrumentation markup language. Retrieved from http://www.nacimiento. com/VIML

Nunnally, J. (1978). Psychometric theory (2nd ed.). USA: McGraw-Hill Publishing Company. (Original work published 1967).

Namlu, A. G. (2003). The effect of learning strategy on computer anxiety. Computers in Human Behavior, 19, 565-578.

O’Neill, K., Singh, G., & O’Donoghue, J. (2004). Implementing e-learning programmes for higher education: A review of the literature. Journal of Information Technology Education, 3, 312-323.

Nanyang Technological University’s OnlineLab. (n.d.). Retrieved from http://www.onlinelab.eee.ntu.edu. sg:8000/ Naps, T. L., Roling, G., Almstrum, V., Dann, W., Fleischer, R., Hundhausen, C., Korhonen, A., Malmi, L., McNally, M., Rodger, S., and Velazquez-Iturbide, J. A. (2003). Exploring the Role of Visualization and Engagement in Computer Science Education, SIGCSE Bulletin 35. Naps, T.L. (2005). JAVÉ: Supporting algorithm animation. IEEE Computer Graphics and Applications, 25(5), 49-55. National Instruments Corporation. (n.d.). Retrieved from http://www.ni.com National University of Singapore’s Vlab. (n.d.). Retrieved from http://vlab.ee.nus.edu.sg NEC Engineering. (1996). Remote conference system. Retrieved from http://www.nec-eng.com/kaigi/index. html (in Japanese) Nichols, M. (2003). A theory for e-learning. Educational Technology & Society, 6(2), 1-10. Niemi, H. (2002). Empowering learners in the virtual university. In H. Niemi, & P. Ruohotie (Eds.), Theoretical understandings for learning in the virtual university. University of Tampere, Research Center for Vocational Education and Training. Novick, L. R. (1990). Representational transfer in problem solving. Psychological Science, 1, 128-132.

0

Object Management Group’s Unified Modeling Language. (n.d.). Retrieved from http://www.omg.org/ OCW. (2004). The MIT OpenCourseWare update: July 2004. Retrieved July 27, 2006, from http://ocw.mit.edu/ OcwWeb/Global/AboutOCW/newsletter.htm ODRL. (2003). Open Digital Rights Language. Retrieved July 27, 2006, from http://odrl.net/ OMG Group. (1993). OMG Unified Modeling Language Specification. Retrieved from http://www.rational.com/ media/uml/post.pdf Oppenheim, A.N. (2001). Questionnaire design (New ed.). New York: Irwin. Organ, O., & Bateman, T. (1991). Organizational behaviour (4th ed.). USA: Irwin. (Original work published 1978). Ouksel, A.M. (1999). A framework for a scalable agent architecture of cooperating heterogeneous knowledge sources. In M. Klusch (Ed.), Intelligent information agents (pp. 100-124). Berlin: Springer-Verlag. Pape, C., & Schmitt, P.H. (1997). Visualizations for proof presentation in theoretical computer science education. In Z. Halim, T. Ottmann, & Z. Razak (Eds.), Proceedings of International Conference on Computers in Education (ICCE ’97) (pp. 229-236). AACE - Association for the Advancement of Computing in Education. Paris, S. G., Cross, D. R., & Lipson, M. Y. (1984). Informed strategies for learning: A program to improve

Compilation of References

children’s reading awareness and comprehension. Journal of Educational Psychology, 7, 1239-1252.

visualization: Programming as a multi-media experience, (pp. 453-480). Cambridge, MA: MIT Press.

Passmore, D.L. (2000). Impediments to adoption of Webbased program delivery among university faculty. ALN Magazine, 4(2). Retrieved May 2004, from http://www. sloan-c.org/publications/magazine/v4n2/passmore.asp

Pham, V. A., & Karmouch, A. (1998). Mobile software agents: An overview. IEEE Communications Magazine, 36(7), 26-37.

Patel, A., & Kinshuk. (1996). Artificial intelligence in learning and assessment — A supportive software for the teaching of accounting. In Proceedings of the 6th EastWest International Conference on Human-Computer Interaction, Moscow (pp. 1-11). Patel, A., & Kinshuk. (1997). Intelligent tutoring tools in a computer integrated learning environment for introductory numeric disciplines. Innovations in Education and Training International Journal, 34(3), 200-207. Patel, A., & Kinshuk. (1997, June 2-6). Intelligent tutoring tools on the Internet — Extending the scope of distance education. In Proceedings of the International Conference on Distance Education (ICDE 97), State College, PA. Retrieved from http://infosys.massey. ac.nz/~kinshuk/papers/icde97.pdf Pea, R. D., & Kurland, D. M. (1984). On the cognitive effects of learning computer programming. New Ideas in Psychology, 2, 137-168. Peck, M.L (1983). Aptitude treatment interaction research has educational value. In Proceedings of selected research paper presentations at the 1983 convention of the Association for educational communicators and technology and sponsored by the research and theory division. (pp. 564-622). New Orleans: Association for Educational Communications & Technology. Petre, M., Baecker, R., & Small, I. (1998). An introduction to software visualization. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software visualization: Programming as a multi-media experience (pp. 3-26). Cambridge, MA: MIT Press. Petre, M., Blackwell, A.F., & Green, T.R.G. (1998). Cognitive questions in software visualization. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software

Piaget, J. (1926). The language and thought of the child. London: Routledge & Kegan Paul. Pintrich, P. R. (2003). A motivational science perspective on the role of student motivation in learning and teaching contexts. Journal of Educational Psychology, 95(4), 667-686. Polycom. (2004). Retrieved from http://www.polycom. com/home Potosky, D. (2002). A field study of computer efficacy beliefs as an outcome of training: The role of computer playfulness, computer knowledge, and performance during training. Computers in Human Behavior, 18, 214-255. Pöyry, P., & Puustjärvi, J. (2003). CUBER: A personalised curriculum builder. In Proceedings of the 3rd IEEE International Conference on Advanced Learning Technologies, Athens, Greece (pp. 326-327). Pöyry, P., Pelto-Aho, K., & Puustjärvi, J. (2002). The role of meta data in the CUBER system. In Proceedings of the Annual Conference of the SAICSIT 2002 (pp. 172-178). Preece, J., Rogers, Y., & Sharp, H. (2002) Interaction design: Beyond human-computer interaction. New York: John Wiley & Sons. Price, B., Baecker, R., & Small, I. (1998). An introduction to software visualization. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software visualization: Programming as a multi-media experience, (pp. 7-27). MIT Press. Price, B.A., Baecker, R., & Small, I. (1993). A principled taxonomy of software visualization. Journal of Visual Languages and Computing, 4(3), 211-266. Ramsden, P. (Ed.). (1988). Context and strategy: Situational influences on learning. In Learning strategies and learning styles. New York: Plenum Press.



Compilation of References

Ratnasamy, S., Ermolinskiy, A., & Shenker, S. (2006). Revisiting IP multicast. In Proceedings of the ACM SIGCOMM 2006 Conference. Redfern, S., & Naughton, N. (2002). Collaborative virtual environments to support communication and community in Internet-based distance education. Journal of Information Technology Education, 1(3), 201-211. Rensselaer Polytechnic Institute’s AIM-Lab. (n.d.). Retrieved from http://nina.ecse.rpi.edu/shur/remote/ Rentroia-Bonito, M. A., & Jorge., J. A. (2004). Motivation to e-learn within organizational settings: What is it and how could it be measured? In Proceedings of 15th IRMA International Conference, New Orleans, LA. IRMA. Retrieved December 2005 from http://citeseer. ist.psu.edu/700830.html Rentroia-Bonito, M. A., Jorge, J.A., & Ghaoui, C. (2006). Motivation to e-learn within organizational settings: An exploratory factor structure. International Journal of Distance Education Technologies, 4(3), 24-35. Rose, R.G. (2001, February 8). Practical use of pilot personality profile. Retrieved February 18, 2003, from http://www.avweb.com/news/aeromed/181606-1.html Rosenberg, M. (2001). Mixing apples and oranges: Quick tips for surviving the interoperability myth. E-Learning Magazine, 2(10), 30-31. Rosenblatt, B., Trippe, B., & Mooney, S. (2001). Digital rights management: Business and technology (1st ed.). New York: John Wiley & Sons. Rößling, G., & Freisleben, B. (2002). ANIMAL: A system for supporting multiple roles in algorithm animation. Journal of Visual Languages and Computing, 13(3), 341-354. Rößling, G., Naps, T., Hall, M., Karavirta, V., Kerren, A., Leska, C., Moreno, A., Oechsle, R., Rodger, S. H., Urquiza-Fuentes, J., & Velázquez-Iturbide, J.A. (2006). Merging interactive visualizations with hypertextbooks and course management. ACM SIGCSE Bulletin – Inroads, 38(4), 166-181.



Rößling, G., Schüer, M., & Freisleben, B. (2000). The ANIMAL algorithm animation tool. In Proceedings of the 5th annual SIGCSE/SIGCUE ITiCSE Conference on Innovation and Technology in Computer Science Education, Helsinki, Finland, (pp. 37-40). New York: ACM Press. Rumetshofer, H., & Wöß, W. (2003). XML-based adaptation framework for psychological-driven e-learning systems. Educational Technology & Society, 6(4). Retrieved July 19, 2006, from http://ifets.ieee.org/periodical/6_3/4.pdf Ryan, S., Scott, B., Freeman, H., & Patel, D. (2000). The virtual university. The Internet and resource-based learning. London: Kogan Page. Ryan, Y. (2002). Emerging indicators of success and failure in borderless higher education. Observatory on Borderless Higher Education. Retrieved March 2006, from http://www.obhe.ac.uk/products/reports/pdf/February2002.pdf Sackett, P. R., & Mullen, E. J. (1993). Beyond formal experimental design: Towards an expanded view of the training evaluation process. Personnel Psychology, 46, 613-627. Safoutin, M.J., Atman, C.J., Adams, R., Rutar, T., Kramlich, J.C., & Fridley, J.L. (2000). A design attribute framework for course planning and learning assessment. IEEE Transactions on Education, 43, 188-199. Sakiyama, T., Ohono, N., Mukunoki, M., & Ikeda, K. (2001). Video stream selection according to lecture context in remote lecture. IEICE, 84-D-II(2), 248-257. (in Japanese) Salas, E., & Cannon-Bowers, J. A. (2001). The science of training: A decade of progress. Annual Review of Psychology, 52, 471-499. Salomon, G., & Perkins, D. N. (1988). Teaching for transfer. Educational Leadership, 46(1), 22-35. Sand, K., & Schoenfelder, J. (1998). Simulation coupled with CBT creating a comprehensive training tool that

Compilation of References

increases transfer. Retrieved February 17, 2003, from www.viats.org/papers/session5a4.htm, 1

Senvid’s summit system: Remote laboratory. (n.d.). Retrieved from https://www.senvid.net/

Sandhu, R. S., Coynek, E. J., Feinsteink, H. L., & Youmank, C. E. (1996). Role-based access control model. IEEE Computer, 29(2), 38-47.

Shavelson, R. J., & Towne, L. (Eds.). (2002). Scientific research in education. Washington, DC: The National Academies Press.

Santally, M. (2003). Individual instruction and distance learning: Application of learning and cognitive styles. Malaysian Journal Of Distance Education, 5(2), 1526.

Shebilske, W.L., Jordan, J.A., Goettl, B.P., & Day, E.A. (1999). Cognitive and social influences in training teams for complex skills. Journal of Experimental Psychology: Applied, 5(3), 227-249.

Santally, M., Govinda, M., & Senteni, A. (2004). Reusable learning object aggregation for e-learning courseware development at the University of Mauritius. International Journal of Instructional Technology and Distance Learning. Retrieved July 19, 2006, from http://www.itdl. org/Journal/Jul_04/article02.htm

Sherry, L. (1996). Issues in distance learning. International Journal of Educational Telecommunications, 1(4), 337-365. Retrieved February 19, 2003, from http://carbon. cudenver.edu/~lsherry/pubs/issues.html

Santhanam, R., & Sein, M. K. (1994). Improving enduser proficiency: Effects of conceptual training and nature of interaction. Information Systems Research, 5(4), 378-399. Sapporo-shi. (2004). Tokeidai at the moment. Retrieved from http://web.city.sapporo.jp/livecamera/ Satyanarayanan, M. (2001). Pervasive computing: Vision and challenges. IEEE Personal Communications, 8(8), 10-17. Scalable Vector Graphics (SVG). (2007). XML Graphics for the Web. Retrieved from http://www.w3.org/Graphics/SVG/ Schar, S.G., & Kruger, H. (2000). Using new learning technologies with multimedia. IEEE Multimedia, 7(3), 40-51. Schneider, D. (2003). Conception and implementation of rich pedagogical scenarios through collaborative portal sites: Clear focus and fuzzy edges. In Proceedings of the International Conference on Open & Online Learning, ICOOL 2003, Mauritius. Retreived December 5, 2004 from http://icool.uom.ac.mu/2003/papers/schneider. php Selvaratnam, S. (2002, November 11). Tool to create e-learning content. Computimes Malaysia.

Shi, Y. C., & Xie, W. K. (2003). The smart classroom: Merging technologies for seamless tele-education. IEEE Pervasive Computing Magazine, 2(2), 25-33. Short, J., Williams, E., & Christie, B. (1976). The social psychology of telecommunications. New York: Wiley. Shuell, T. J. (1986). Cognitive conceptions of learning. Review of Educational Research, 56(4), 411-436. Sidman, M. (1960). Tactics of scientific research. New York: Basic Books. Silva, A., Silva, M. M., & Delgado, J. (1998). AgentSpace: A next-generation mobile agent system. Lecture Notes in Computer Science, 1477, 148-159. Silva, L. M., Simões, P., Soares, G., Martins, P., Batista, V., Renato, C., et al. (1999). JAMES: A platform of mobile agents for the management of telecommunication networks. In Proceedings of the 3rd International Workshop on Intelligent Agents for Telecommunication Applications, Vol. 1669, (pp. 77-95). Stockholm, Sweden. Sime, M. E., Green, T. R. G., & Guest, D. J. (1973). Psychological evaluation of two conditional constructions used in computer languages. International Journal of Man-Machine Studies, 5, 105-113. Simmons, R., & Apfelbaum, D. (2001). A task description language for robot control. In Proceedings Conference on Intelligent Robotics and Systems, New York (Vol. 1, No. 10, pp. 138-147).



Compilation of References

Simon, S. (2002). A model for software configuration in ubiquitous computing environments. In Proceedings of Pervasive. LNCS 2414, Zürich (Vol. 1, No. 7, pp. 181-194). Simon, S. J., & Werner, J. M. (1996). Computer training through behavior modeling, self-paced, and instructional approaches: A field experiment. Journal of Applied Psychology, 81(6), 648-659. Simon, S. J., Grover, V., Teng, J. T. C., & Whitcomb, K. (1996). The relationship of information system training methods and cognitive ability to end-user satisfaction, comprehension, and skill transfer: A longitudinal field study. Information Systems Research, 7(4), 466-490. Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. New York: Appleton-Century Company, Incorporated. Small, R.L., Lakowske , S.D., Breese, J., & Callejo, G. (1999). A future direction in pilot training. Retrieved February 17, 2003, from www.viats.org/papers/session5d2.htm, 1-2 Sohn, Y. W., & Doane, S. M. (1997). Cognitive constraints on computer problem-solving skills. Journal of Experimental Psychology: Applied, 3, 288-312. Soloway, E., & Pryer, A. (1996). The next generation in human-computer interaction. Communications of the ACM, 39(4), 16-18. SONY. (2004). Video conference system PCS-1. Retrieved from http://bssc.sel.sony.com/Professional/markets/videoconferencing/overview.html Special Issue. (2004). Design-based research methods for studying learning in context. Educational Psychologist, 39(4), 199-260. Stasko, J., & Lawrence, A. (1998). Empirically assessing algorithm animations as learning tools. In J. Stasko, J. Domingue, M.H. Brown, & B.A. Price (Eds.), Software visualization: Programming as a multi-media experience, (pp. 417-419). Cambridge, MA: MIT Press. Stasko, J. T. (1996). Using student-built algorithm animations as learning aids (Tech. Rep. GIT-GVU-96-19).



Graphics, Visualization, and Usability Center, Georgia Institute of Technology. Steeples, C., & Jones, C. (2002). Networked learning: Perspectives and issues. London, UK: Springer. Stern, L., & Naish, L. (2002). Animating recursive algorithms. Interactive Multimedia Electronic Journal of Computer Enhanced Learning, 4(2). Stern, L., & Naish, L. (2002). Visual representation for recursive algorithms. In Proceedings of the 33rd Annual SIGCSE Technical Symposium on Computer Science Education, (pp. 196-200). Association for Computing Machinery. Sternberg, R. J., & Lyon, G. R. (2002). Making a difference to education: Will psychology pass up the chance? Monitor on Psychology, 33, 76. Retrieved April 6, 2004, from http://www.apa.org/monitor/julaug02/difference. html Stojanovic, L., Staab, S., & Studer, R. (2001). E-learning based on the Semantic Web. In Proceedings of WebNet2001 — World Conference on the WWW and Internet, Orlando, FL. Stokes, S. (2000, July 24). Learning space enlivens online training. Infoworld. Straub, D. W. (1994). The effect of culture on IT diffusion: E-mail and FAX in Japan and the U.S. Information Systems Research, 5(1), 23-47. Straub, D., Keil, M., & Brenner, W. (1997). Testing the technology acceptance model across cultures: A three country study. Information and Management, 33, 1-11. Strijker, A. (2001). Using metadata for re-using material and providing user support tools. In Proceedings of ED-MEDIA 2001, Tampere, Finland. Sudmann, N. P., & Johansen, D. (2001). Supporting mobile agent applications using wrappers. In Proceedings of the 12th International Workshop on Database and Expert Systems Applications (pp. 689-695). Munich, Germany.

Compilation of References

Svastisinha, R. W. (1999). Wahhn: Web-based design. Wind and human comfort for Thailand. Unpublished doctoral dissertation, University of Southern California.

The House of Councilors. (2005). The House of Councilors Internet TV. Retrieved from http://www.webtv. sangiin.go.jp/webtv/index.php (in Japanese)

Swezey, R. W., & Llaneras, R. E. (1997). Models in training and instruction. In G. Salvendy (Ed.), Handbook of human factors and ergonomics (pp. 514-577). New York: Wiley.

The House of Representatives. (1999). The House of Representatives Internet TV. Retrieved from http://www. shugiintv.go.jp/top.cfm

Takasugi, K. (2001). Adaptive system for service continuity in a mobile environment. In Proceedings of IEEE APCC, Tokyo, Japan (Vol. 1, No. 9, pp. 75-83). Takasugi, K. (2003). Seamless service platform for following a user’s movement in a dynamic network environment. In Proceedings of PerCom’03 (Vol. 1, No. 8, pp. 125-132). Tan, S.S., & Ng, F. (2006). A problem-based learning approach to entrepreneur-ship education. Education & Training, 48(6), 416 Tanudjaja, F., & Mui, L. (2002). Persona: A contextualized and personalized Web search. In Proceedings of the 35th Annual Hawaii International Conference on System Sciences (HICSS’02) (Vol. 3, pp. 53). Taylor, J. C. (2001, June). Fifth generation distance education. Higher Education Series (Report No.40). Canberra: DETYA. Retrieved March 2007, from http://www. dest.gov.au/archive/highered/hes/hes40/hes40.pdf Taylor, S., & Todd, P. (1995). Understanding information technology usage: A test of competing model. Information Systems Research, 6(23), 144-176. Teare, R., Davies, D., & Sandelands, E. (1999). The virtual university — An action paradigm and process for workplace learning. Cassell. Texas Instrument. (n.d.). Retrieved from http://www. ti.com.tw Thaler, D., & Handley, M. (2000). On the aggregatability of multicast forwarding state. In INFOCOM 2000. Proceedings of the Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Vol. 3, pp. 1654-1663). IEEE.

The industry model matrix. (n.d.). The British Computer Society. Retrieved September 30, 2005, from http://www. bcs.org/bcs/products/corporate/industrymodel/about/ matrix.htm The Ministry of Public Management. (2004). Home affairs, posts, and telecommunication. Retrieved from http://www.soumu.go.jp/media/index.html The Semantic Web. (n.d.). Retrieved September 30, 2005, from http://www.w3.org/2001/sw/ Thomson, J.R., & Cooke J. (2000). Generating instructional hypermedia with APHID. In Proceedings of the Eleventh ACM Conference on Hypertext and Hypermedia, San Antonio, Texas, USA (pp. 248-249). Thorndike, R. L. (1949). Personnel selection: Test and measurement techniques. New York: John Wiley & Sons. Tomlinson, C.A. (1999, September). Mapping a route towards differentiated instruction. Educational Leadership. Torkzadeh, G., & Van Dyke, T. P. (2002). Effects of training on Internet self-efficacy and computer user attitudes. Computers in Human Behavior, 18, 479-494. Towne, L., & Hilton, M. (Eds.). (2004). Implementing randomized field trials in education. Washington, DC: The National Academies Press. Trapp, R. (1997, September 3). Internet users ‘wide open” to fraud. Independent, p. C39. Retrieved February 19, 2003, from http://www.elibrary.com/. Tsang, A. C. W., & Chan, N. (2004). An online problembased model for the learning of Java. Journal of Electronic Commerce in Organizations, 2(2), 55-64.



Compilation of References

Turoff, M., Hiltz, S. R., Bieber, M., Fjermestad, J., & Rana, A. (1999). Structures in computer mediated group communications. Journal of Computer Mediated Communication, 4(4). Turpeinen, M. (2000). Customizing news content for individuals and communities. In Acta Polytechnica Scandinavica. Mathematics and computing series no. 103. Doctoral thesis, Helsinki University of Technology. Tusch, R., Böszörményi, L., Goldschmidt, B., Hellwagner, H., & Schojer, P. (2004). Offensive and defensive adaptation in distributed multimedia systems. ComSIS, 1(1). U.S. News. (2003). E-learning: Online graduate degrees. Retrieved November 18, 2003, from http://www.usnews. com/ usnews/edu/elearning/directory/gradonline.htm U.S. News. (2007). E-learning guide. Retrieved January 10, 2007, from http://www.usnews.com/usnews/edu/elearning/ tables/mba_reg.htm United States General Accounting Office. (1999). Aviation safety: Research supports limited use of personal computer aviation training devices for pilots (GAO Report No. RCED-99-143). Washington, DC: U.S. Government Printing Office. Retrieved on February 19, 2003, from http://www2.faa.gov/nsp/nsp/GAO_PCATD. txt FAA 1999. University Graduate Center’s LAB-on-WEB. (n.d.). Retrieved from http://www.lab-on-web.com/ Vahidov, R., & Fazlollahi, R. (2003). A multi-agent DSS for supporting e-commerce decisions. Journal of Computer Information Systems, Winter, 87-94. Vandalore, B., Feng, W., Jain, R., & Fahmy, S. (2001). A survey of application layer techniques for adaptive streaming of multimedia. Real Time Imaging, 7(3), 221-235. Veenman, M. V. J., Prins, F. J., & Elshout, J. J. (2002). Initial inductive learning in a complex computer simulated environment: The role of metacognitive skills and intellectual ability. Computers in Human Behavior, 18, 327-241.



Ventatesh, V., Morris, M., Davis, G., & Davis, F. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3). Visual Nexus. (2004). Visual online conference system. Retrieved from http://www.visualnexus.com/en/index. html Vrasidas, C. (2002). A systematic approach for designing hypermedia environments for teaching and learning. International Journal of Instructional Media, 29(1). Retrieved from http://www.cait.org/vrasidas/hypermedia.pdf VTEL. (2004). Retrieved from http://www.vtel.com/ Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. W3C RDF model and syntax specification. (n.d.). Retrieved September 30, 2005, from http://www.w3.org/ TR/1999/REC-rdf-syntax-19990222 W3C’s SOAP. (n.d.). Retrieved from http://www.w3.org/ TR/soap/ Wakahara, T. (1998). Configuration and characteristics of distance learning system over ATM-PVC Network. IEICE, J81-B-I(8), 494-506. (in Japanese). Walker, J. (1992). Human resources strategy. McGrawHill Series in Management. USA: McGraw-Hill. (Original work published 1980). Walker, S. (2003). Distance education learning environments research: A short history of a new direction in psychosocial learning environments (Technical Report). TCC. Wang, H-W. (2006). The implementation of an asynchronous Web-based gaming & simulation environment and its effects on financial engineering education. Journal of American Academy of Business, 9(2), 360-369. Watkins, C. L., & Slocum, T. A. (2004). The components of direct instruction. Journal of Direct Instruction, 3(2), 75-110.

Compilation of References

Watson, J., Ahmed, P.K., & Hardaker, G. (2007). Creating domain independent adaptive e-learning systems using the sharable content object reference model, Campus - Wide Information Systems, 24(1), 45. WBT Systems. (n.d.). Retrieved from http://www.wbtsystems.com Weiser, M. (1991). The computer for the twenty-first century. Scientific American, 265(3), 94-104. Wentling, T., Waight, C., La Fleur, J., Wang, C., & Kanfer, A. (2000). E-learning: A review of literature (Technical Report). University of Illinois (Urbana-Champaign). Wentling, T.L., Waight, C., Strazzo, D., File, J., La Fleur, J., & Kanfer, A. (2000, September). The future of e-learning: A corporate and an academic perspective. NCSA Knowledge and Learning Systems Group. Whitley, J., & King, A. (2001). Animation of insertion sort. Retrieved from http://www.cs.duke.edu/courses/ spring01/ cps049s/students/ack11/jawaasorting.html Wiedenbeck, S., & Zila, P. L. (1997). Hands-on practice in learning to use software: A comparison of exercise, exploration, and combined formats. ACM Transactions on Computer-Human Interaction, 4(2), 169-196. Wilhelm, R., Müldner, T., & Seidel, R. (2001). Algorithm explanation: Visualizing abstract states and invariants. Springer LNCS issue on Software Visualization. In S. Diehl (Ed.), Revised Lectures on Software Visualization, International Seminar (LNCS 2269, pp. 381-394). Williams, M. R. (2006, September 19). Study says teachers are poorly trained. Knight Ridder Tribune Business News, p. 1 Wittrock, M. C. (1978). The cognitive movement in instruction. Educational Psychologist, 13, 15-29. Witzel, M. (2005, October 7). A new era for digital learning. Financial Times, p. 2 Wolf, C. (2002). iWeaver: Towards an interactive Webbased adaptive learning environment to address individual learning styles. European Journal of Distance

Learning. Retrieved July 19, 2006, from http://www. eurodl.org/materials/contrib/2002/2HTML/iWeaver Wu, D., Hou, Y., Zhu, W., Zhang, Y., & Peha, J. (2001). Streaming video over the Internet: Approaches and directions. IEEE Trans. Circuits Syst. Video Technol., 11(3), 282-300. Wu, H. T., Hsiao, T. C., Chen, C. L., Su, C. M., Su, J. C., & Jiang, J. C. (2001). An integrated teaching and learning DSP lab. system. Journal of Science and Technology, 10(1), 29-36. XrML. (2002). Extensible Rights Markup Language specification. Retrieved July 27, 2006, from http://www. xrml.org/ Yamaguchi, T., Sakano, T., Fujii, T., Ando, Y., & Kitamura, M. (2001). Design of medical tele-consultation support system using super high definition imaging system. IEICE, J84-D-II(6), 1203-1212. (in Japanese) Yang, J. T., & Tsai, C. Y. (2003). An implementation of SCORM-compliant learning content management system-content repository management system. In Proceedings of the ICALT2003, Athens, Greece. Yi, M. Y., & Davis, F. D. (2001). Improving computer training effectiveness for decision technologies: Behavior modeling and retention enhancement. Decision Sciences, 32(3), 521-544. Yoo, Y., & Alavi, M. (2001). Media and group cohesion: Relative influences on social presence, task participation, and group consensus. MIS Quarterly, 25(3), 371-390. Yoshida, T. & Tada, K. & Hangai, S. (2003). A keyword accessible lecture video player and its evaluation. In Proceedings of IEEE ITRE2003, International Conference on Information Technology, Research, and Education (pp. 610-614). Yoshida, T., Tada, K., Hamamoto, T., & Hangai, S. (2002). A keyword accessible AV-stream player. In Proceedings of ISAS SCI2002 IVRCIA, 16, 100-104. You, S., Wang, T., Eagleson, R., Meng, C., & Zhang, Q. (2001). A low-cost Internet-based telerobotic system for



Compilation of References

access to remote laboratories. Artificial Intelligence in Engineering, 15(3), 265-279. Young, J. D. (1996). The effect of self-regulated learning strategies on performance in learner controlled computerbased instruction. Educational Technology Research and Development, 44, 17-27. Zhang, B. K., & Seo, Y. W. (2001). Personalized Webdocument filtering using reinforcement learning. Applied Artificial Intelligence, 15(7), 665-685.



Zhou, Y., & Evens, M. (1999). A practical student model in an intelligent tutoring system. In Proceedings of the 11th International Conference on Tools with Artificial Intelligence, Chicago, IL (pp. 13-18). Zimmerman, B. J. (1994). Dimensions of academic selfregulation: A conceptual framework for education. In D. H. Schunk, & B. J. Zimmerman (Eds.), Self-regulation of learning and performance (pp. 3-21). Hillsdale, NJ: Erlbaum.

329

About the Contributors

Paulo Carvalho graduated in 1991 and received his PhD degree in computer science from the University of Kent at Canterbury, Canterbury, United Kingdom, in 1997. He is currently assistant professor of computer communications, Department of Informatics, at the University of Minho, Portugal. His main research interests include broadband technologies, multiservice networks, mobile networks, and data traffic analysis, characterization and modelling. Charlie C. Chen is an assistant professor in the Department of Computer Information Systems at Appalachian State University. He received his PhD in management information systems from Claremont Graduate University in 2003. He has authored more than 30 referred articles and proceedings, presented at many professional conferences and venues. Dr. Chen has published in journals such as Communications of the AIS, Journal of Knowledge Management Research Practice, and Journal of Information Systems Education. Dr. Chen is a project management professional (PMP) certified by the Project Management Institute. Dr. Chen is working on improving information system solutions infrastructural, managerial, and operational perspectives. His current main research areas are online learning, mobile commerce, and supply chain technology. Mohammed Chowdhury is an adjunct faculty member at the University of Dallas. He served as assistant professor of business administration in Jahangirnagar University and Chittagong University. He served as manager in the IT industry of Bangladesh. He has extensive experience in consulting, training and re-engineering at various SMEs in Bangladesh. He has several research publications in academic and professional journals and served as a member of editorial review board for the Journal of Business Administration. He is a member of Sigma Iota Epsilon. Chris Colquitt is an aviation and aerospace professional with over 28 years of experience. After earning a BS in aerospace engineering from the University of Texas at Austin he served as a pilot and officer in the United States Navy. After seven years in the Navy he joined American Airlines and flew their flagships all over the globe. Later he accepted a position as a systems matter expert for the U.S. Air Force KC-10 Pilot training program with Raytheon and a systems engineer for the U.S. Navy’s new F18 simulator development. While earning an MBA from the University of Dallas in Irving he flew as a pilot for Piedmont and trained and evaluated professional pilots at CAE SimuFlite. Currently he is on contract with NASA as a research pilot and aerospace engineer.

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

About the Contributors

Gennaro Costagliola is currently professor at the University of Salerno. His research interests include programming languages, visual languages, parsing technologies, multimedia databases, Web technologies and e-learning. He received the Laurea degree in computer science from the University of Salerno, Italy in 1987, and an MS degree in computer science from the University of Pittsburgh in 1991. He was guest co-editor of the February 2002 Special Issue of the Journal of Visual Languages and Computing on Querying Multiple Data Sources. He is a member of ACM, IEEE, and IEEE Computer Society. Sérgio Deusdado received his MSc degree in computer science from the University of Minho, Portugal, in 2002. Since 2004, he has been a PhD student, developing bioinformatics investigation, in the Department of Informatics at the University of Minho. He is currently a lecturer in computer science at the Polytechnic Institute of Bragança in Portugal. His research interests include bioinformatics, information theory, internetworking, groupware and e-learning. Filomena Ferrucci received the Laurea degree in computer science (cum laude) from the University of Salerno (Italy) in 1990. In 1995 she received the PhD in applied mathematics and computer science at the University of Naples (Italy). From 1995 to 2001 she has been a research associate at the University of Salerno where she is associate professor in computer science since November 2001. She was program co-chair of the 14th International Conference on Software Engineering and Knowledge Engineering and guest editor of the special issue of the International Journal of Software Engineering and Knowledge Engineering dedicated to a selection of the best papers of the conference. She has served as program committee member for several international conferences. Her research interests are in the fields of human-computer interaction, e-learning, and software engineering. She is co-author of about 60 papers published on international journals and proceedings of international conferences. Claude Ghaoui is a senior lecturer in computer systems (since 1995) at the School of Computing and Mathematical Sciences, Liverpool John Moores University, UK. Her research insterest and expertise are mainly in human computer interaction, multimedia technology and their application in education. From 1991 to 1995, she worked at the Computer Science Departament of Liverpool University as a pubishing director of an active R&D group of 25 staff. Prior to that, she worked at IBM and at Kuwait University. She passionately promotes innovative IT for the provision of flexible learning. She has numerous publications and has organized a number of international workshops and sessions since 1994. She is a UK correspondent for EUROMICRO (1998-2003), and has served on program committees for several international HCI/multimedia conferences. She is an advisor (since 2000) for University. Net, which promotes online learning, research and innovation. She co-edited several books: Medical Multimedia, Usability Evaluation of Online Learning programs (2003) and e-Educational applications: human Factors and Innovativeapproahces (2004), Knowledge-Based Virtual Education: User-Centred Paradigms (Studies in Fuzziness and Soft Computing) and Encyclopedia Of Human Computer Interaction (both in 2005). Seiichiro Hangai received a BS and PhD from Tokyo University of Science, in 1975 and 1981, respectively. Hangai has served as research assistant in 1981, associate professor in 1991 and professor in 2001 at the same university. He is also director of Center of Information Science and Education, and professor of Master of Intellectual Property, Tokyo University of Science, and was a visiting scholar at the CSLI,

0

About the Contributors

Stanford University from 1996 to 1997. He is engaged in the research of speech, image and security signal processing. He is a member of IEEE, IEICE Japan, and ITE Japan. Albert L. Harris is a professor in the Department of Computer Information Systems at Appalachian State University and editor-in-chief of the Journal of Information Systems Education. He received his PhD in MIS from Georgia State University. He is a member of the board of directors of the International Association of Information Management and the Education Special Interest Group of AITP. His research is focused on IS education, IS ethics, and international issues in IS. He has more than 60 publications in journals, books, and proceedings, including the Communications of the AIS, Information & Management, Journal of Information Systems Education, Journal of Computer Information Systems, International Journal of Technology and Human Interaction, Knowledge Management Research & Practice, International Journal of Management, and International Journal of Computer Applications in Technology. J. A. Jorge is associate professor at Instituto Superior Técnico (IST/UTL) Lisboa, Portugal and leads the Intelligent Multimodal Interfaces Group (http://immi.inesc-id.pt) at INESC-ID. He received PhD and MSc degrees in computer science from Rensselaer (New York), co-chaired 20 international conferences, including EUROGRAPHICS’98, served on over 90 international conference committees and published over 110 internationally refereed papers. He is editor-in-chief of Computers & Graphics Journal and serves on editorial boards of Computer Graphics Forum and IJDET Journals. He is affiliated with ACM/ SIGGRAPH, IEEE Computer Society (SM’2000) and IFIP TC13. His research interests are distance learning, multimodal user interfaces, and visual languages. Andreas Kerren is currently an associate professor in computer science at the School of Mathematics and Systems Engineering of Växjö University, Sweden. He was involved in various successful research projects related to computer science education, for example, in the DFG project “Generation of Interactive Multimedia Visualizations and Animations for Learning Software in Compiler Design.” Dr. Kerren is a member of several program and organizing committees. He has served as reviewer for several international journals and distinguished conferences. His main research interests lie in the areas of software visualization, information visualization, software engineering, computer science education, human-computer interaction, and programming languages. Iwona Miliszewska is a senior lecturer in computer science at Victoria University, Melbourne, Australia. She has led and participated in research projects involving distance education, transnational education, collaborative education systems, effective teaching methods, lifelong learning and women in computer science, and has published in these areas. Iwona is currently leading a grant-funded research project aimed at testing and refining a multidimensional model for effective transnational education programs. Tomasz Müldner is a professor of computer science at Acadia University. He received his PhD in mathematics from the Polish Academy of Sciences in Warsaw, Poland, in 1975. He published several books and over 70 papers. He is the recipient of numerous teaching awards, including the prestigious Acadia University Alumni Excellence in Teaching Award in 1996. His research interests include algorithm explanation, Web site internationalization and security of XML.



About the Contributors

Giuseppe Polese is associate professor in the Department of Mathematics and Computer Science at the University of Salerno, Italy. His research interests include visual languages, multimedia databases, e-learning, and multimedia software engineering. He received the Laurea degree in computer science from the University of Salerno, an MS in computer science from the University of Pittsburgh, and a PhD in computer science and applied mathematics from the University of Salerno. Mahesh S. Raisinghani is an associate professor at TWU School of Management’s Executive MBA program. Dr. Raisinghani was the recipient of the Presidential Award and the King Haggar Award for excellence in teaching, research and service. His previous publications have appeared in IEEE Transactions on Engineering Management, Information and Management, Information Resources Management Journal, Information Strategy: An Executive’s Journal, Journal of Global IT Management, Journal of Digital Information, Journal of E-Commerce Research, Journal of IT Cases and Applications and International Journal of E-Business Research, among others. He serves on the editorial review board for leading information systems publications and is included in the millennium edition of Who’s Who in the World, Who’s Who Among America’s Teachers and Who’s Who in Information Technology. M. A. Rentroia-Bonito received a MSc degree in industrial and labour relations from Cornell University, Ithaca, NY, in 1993; a MBA degree from Catholic University “Andres Bello,” and a bachelor’s degree in systems engineering from IUPLCM, both in Caracas, Venezuela. During the last 20 years, she has worked as a manager, consultant and trainer, mainly within multinational companies. She is currently pursuing her PhD degree in human-computer interaction in the Computer Science Department at Instituto Superior Técnico (IST/UTL). Her research interests are: situated usability, evaluation for e-learning, and e-inclusion. Giuseppe Scanniello received the Laurea degree in computer science from the University of Salerno (Italy) in 2001, where in 2003 he also received the PhD in computer science. In 2006 he joined the Department of Mathematics and Computer Science of the University of Basilicata, Potenza, Italy, where he is currently an assistant professor. His research interests include reverse engineering, reengineering, workflow automation, migration of legacy systems, wrapping, integration, e-learning, cooperative supports for software engineering, and visual languages. He is a member of IEEE Computer Society. Elhadi Shakshuki is an associate professor and he is currently the graduate program coordinator in the Jodrey School of Computer Science at Acadia University, Canada. He is the founder and the head of the Cooperative Intelligent Distributed Systems Group at the Computer Science Department, Acadia University. He received the BSc degree in computer engineering in 1984 from El-Fateh University, and the MASc and PhD degrees in systems design engineering in 1994 and 2000, respectively, from the University of Waterloo, Canada. He manages several research projects in his research expertise in the area of intelligent agent technology and its applications. Rong-An Shang is an associate professor in the Department of Business Administration at Soochow University, Taiwan. He received his MS degree from National Chioao Tung University, Hsinchu, Taiwan, and a PhD degree from National Taiwan University in 1997. His current research interests include electronic commerce, IS adoption and implementation, and computer-mediated communication. His



About the Contributors

work has been published in Information & Management, Journal of Business Ethics, Internet Research, International Journal of Distance Education Technologies, as well as some Chinese journals and international conference proceedings. Takahiro Yoshida received BS, MS and PhD from Tokyo University of Science, in 1999, 2001, and 2004, respectively. From April 2004, he is an assistant professor at Department of Electrical Engineering, Tokyo University of Science. His research interests include speech signal processing, image processing, e-learning system and electrostatic discharge. He is a member of IEEE, IEICE Japan, ITE Japan, IPS Japan, and IEJ.



334

Index A abstract algorithm model (AAM) 66 abstract data type (ADT) 60 abstract iterator implementation model (AIIM) 73 active learning 33 adaptation, concept of 235 advanced distributed learning (ADL) 167 agent-based communication, over Internet, 84 agent-based software framework 90 agent environment (AE) 272 aggregation portal 196 airline transport rating (ATP) 125 algorithm, defnition of 59 algorithm animation (AA) 60 algorithm explanation (AE) 63 algorithms in action (AIA) 61, 64 algorithm visualization (AV) 60 a new interactive modeler for animations in lectures (ANIMAL) 61 any source multicast (ASM) 8 apparatus virtual user interface (App-VUI) 171 aptitude treatment interaction (ATI) 235 artificial intelligence (AI) 98 asynchronous BM (ABM) 20 AT&T Broadband 128 authentication manager (AM) 255 aviation education reinforcement option (AERO) 127 aviation industry 124 Aviation Industry CBT Committee (AICC) 167

click-through-rate (CTR) 44 collaborative education model 84 computer-assisted instruction (CAI) 182 computer-based training (CBT) 125 concept keyboards (CKs) 62 connection manager (CM) 255 constructs efficacy (EFF) 138 content management system (CMS) 166 content repository management system (CRMS) 284, 285 continuous professional development (CPD) 96, 108 cost-per-click (CPC) 44 curriculum vitae (CV) 106 cyber-terrorism, defined 129

D data agent (DA) 272 decomposed theory of planned behavior (DTPB) 132 digital rights management (DRM) 284, 285, 286 digital signal processing (DSP) 181 digital video (DV) camera 112 discretionary access control (DAC) 288 Dunn and Dunn learning-style model 232 Dunn and Dunn learning styles 232

E

behavior modeling (BM) 14, 15 behavior modeling, and knowledge transfer 16 Boolean model 194, 200 business aviation pilots 124

e-learning, challenges 49 e-learning courses 147 e-learning environment 197 e-learning systems 32 e-learning systems, overview 166 ease of use (EU) 138 execution/code agent (EA) 272 extended markup language (XML) 271

C

F

CAE SimuFlite 124 case-based teaching 31

face-to-face (F2F) learning 15, 32 face-to-face BM (FBM) 20

B

Copyright © 2009, IGI Global, distributing in print or electronic forms without written permission of IGI Global is prohibited.

Index

facilitating conditions resources (FCR) 138 fed-Education mode 89 federated model 84 federation dictionary 89 Flight Safety International (FSI) 126 free ride multicasting (FRM) 11 fuzzy queries 196

G graph theory, and RDF 289

H holistic technology-based training models (HTBT) 127

I IEEE Learning Technology Standards Committee (LTSC) 167 IMS Global Learning Consortium (IMS) 167 incremental probabilistic action modeling prediction method (IPAM) 104 industry structure model (ISM) 104 information and communications technology (ICT) 3, 285 information technology (IT) 252 intellectual property rights (IPR) 284, 286 intelligent tutoring systems (ITS) 99 interactive data structure visualization (IDSV) 61 Internet-based learning systems 165 Internet protocol (IP) 184

J Java applet 207 Java Media Framework (JMF) 3 Java native interface (JNI) 187 Java virtual machine (JVM) 185 Jena API 298

K Knowledge Far-Transfer (KFT) 19 Knowledge Near-Transfer (KNT) 18 knowledge transfer 17 knowledge transfer theories 18

L laboratory learning objects (LLO) 173 lateral transfer 18 leaning content management system (LCMS) 301 leaning management system (LMS) 301

learning-objects rights graph (LRGr) 295 learning activity diagrams (LAD) 148 learning management system (LMS) 166 learning object metadata (LOM) 103 learning objects (LOs) 166 LearningSpace™ software 129 learning style inventory (LSI) 233 learning styles questionnaire (LSQ) 233 least-squares deconvolution (LSD) 24 lecture video player/maker system 111 lecture videos 111 lecture videos, desired style of 121 lecture videos, quality of 112 location management agent (LMA) 272

M mandatory access control (MAC) 288 metadata, varieties of 199 MIT Open CourseWare (OCW) 285 mobile agent (MA) 272 mobile agent execution environment (MAEE) 184 mobile distance learning (MDL) 269 mobile Web-based learning 269 motivation-to-e-learn variable 49, 50 multimedia multicast conferences (MMC) 2 multimedia software engineering (MSE) 153, 161 MyProgram, explanation 210

N National Business Aviation Association (NBAA) 125 navigation agent (VA) 272 near transfer vs. far transfer 17 network capability agent (NCA) 272 non-case discussion grade (NCDG) 42

O ONES system 194 one stop e-learning portal (ONES) 196 online asynchronous learning (OAL) 30 open digital rights language(ODRL) 286 Open Universities Australia (OUA) 87

P perceived usefulness (PU) 133, 138 personalization learning (PL) 96 personalized CPD learning portal (PersonalizedCPD) 96 positive transfer vs. negative transfer 17





problem-solving learning 34 program animation (PA) 60

Q QoS adaptation 1 quality of service (QoS) 1

R RDF format, encoding of learning objects 298 RDF instance generator (RIG) 295 real-time data exchange (RTDX) 187 real-time transport protocol (RTP) 3 Remotely Accessible Laboratory (REAL) 167 resource description framework (RDF) graph 284, 285 reusable learning objects (RLOs) 166 rights expression language (REL) 284, 286, 287 rights instance processor (RIP) 295 rights validation module (RVM) 295 role-based access control (RBAC) 288 role-based access control (RBAC) models 285 round-trip time (RTT) 5

S scalable vector graphics (SVG) 156 selected, linear and definite (SLD) 289 self-consistent learning content objects (SCLO) 151 self-consistent learning object (SCLO) 149 semantic metadata 200 service agent (SA) 272 SHALEX agent 76 sharable content object reference model (SCORM) 285 shared object manager (SOM) 255 Simfinity™ technology 124 software engineering (SE) 161 software visualization (SV) 60 Statistical Program for Social Sciences (SPSS) software 133 structured hypermedia algorithm explanation (SHALEX) 59 structured hypermedia algorithm explanation (SHALEX) system 58 structure of the learning outcome (SOLO) 127 synchronized multimedia integration language (SMIL) 271 synchronous BM (SBM) 20 system for e-learning activity management (SEAMAN) 148, 155

T task-oriented mobile distance learning 271 task agent (KA) 272 technology-supported learning 49 technology acceptance model (TAM) 132 terminal agent (TA) 272 test maker language (TML) 149, 151, 155 theory of planned behavior (TPB) 132 tutoring system, seven stages of 210

U unified modeling language (UML) 148, 168 UNIX™ text editor 215 usage request graph (URGr) 295 user agent (UA) 272 user document database agent (UDDA) 272 user interface (UI) 255

V value-adding (process) system 130 vector model 201 vertical transfer 18 video I/O manager (VI/OM) 255 video quality manager (VQM) 255 video stream manager (VSM) 255 virtual digital signal processing laboratory (VDSPL) 180 virtual universities 194 virtual universities, information retrieval in 194 virtual universities, studying in 197 visual languages (VL) 147, 150, 161

W Web-based tutor 207 WWW conference manager (WCM) 255 WWW conference server 254 WWW conference system 253 WWW conference system, proposed 253