Get Complete Project Material File(s) Now! »
EVOLUTION OF E-EDUCATION
The history of E-Education is also a history of communication technologies. As new communication technologies have been developed, they have joined the repertoire of the distance educator. Generally, each of the emerging educational delivery technologies has been incorporated into different e-Education systems, resulting in a total multimedia-based e-Educational system comprised of various generations of distance technology and media. In other words, the different technologies and media have complemented and supported each other, rather than replaced existing ones. Historically, e-Education operations have evolved through the following four generations: first, the Correspondence Model based on print technology; second, the Multi-media Model based on print, audio and video technologies; third, the Tele-learning Model, based on applications of telecommunications technologies to provide opportunities for synchronous communication; and fourth, the Flexible Learning Model based on online delivery via the internet. Although the latter approach is still gaining momentum, as we stride into the new millennium, there is already emerging the fifth generation of e-Education based on the further exploitation of new technologies. The fifth generation has the potential to decrease significantly the cost of online tuition and thereby increase significantly access to education and training opportunities on a global scale. Through the application of automated response systems, which entail the use of software that can scan the text of an incoming electronic message and respond intelligently- without human intervention. In fact, the fifth generation of e-Education is a derivation of the fourth generation, which aims to capitalize on the features of the Internet and the Web. To place the fifth generation Intelligent Flexible Learning Model into a meaningful conceptual framework, it is first worth reviewing briefly certain features of the previous four generations of e-Education. Some of the characteristics of the various models of e-Education that are relevant to the quality of teaching and learning are summarized in Table 2-1, along with an indicator of institutional variable costs (Taylor et al. 1993). In traditional e-Education delivery, the distribution of packages of self-instructional materials (audiotapes, videotapes, etc) is a variable cost, which varies in direct proportion to the number of learners enrolled. Internet-based delivery, however, changes significantly the institutional costs associated with learners gaining access to learning experiences. For example, a key consideration for the fifth generation is the use of automated response systems to reduce the variable cost of computer-mediated communication (CMC), which in the fourth generation is quite resource- intensive.
ADVANTAGES & DISADVANTAGES
The chief advantages of E-Education programs are that learners can learn at their convenience thus accommodating work and personal life and that it can be accessed by those who do not live near or who cannot attend traditional training centers and universities. This is tempered, however, by some of the costs and personal motivation needed to complete programs. For faculty, teaching at a distance requires a large shift in what is normally performed from being just a teacher to being a combination of facilitator, coach, and mentor. Last-minute preparation in isolation cannot happen since one needs to work with a team of professionals. Typically, teaching at a distance requires more time and faculty workload (Billings 1997). Cravener (1999) found in her review of 185 articles that having learners at a distance increased faculty time demands when compared with the classroom courses. For example, in a graduate epidemiology course, administrators complained of the number of e-mails and feedback needed to make learners feel less isolated and supported (Rose et al. 2000).
In e-Education, the learner is usually isolated. The motivational factors arising from the contact and competition with other learners are absent. The learner also lacks the immediate support of a teacher who is present and able to motivate and, if necessary, give attention to actual needs and difficulties that crop up during study. Distant learners and their teachers often have little in common in terms of background and day-to-day experiences and therefore, it takes longer for learner-teacher rapport to develop. Without face-to-face contact distant learners may feel ill at ease with their teacher as an “individual” and uncomfortable with their learning situation. In e-Education settings, technology is typically the conduit through which information and communication flow. Until the teacher and learners become comfortable with the technical delivery system, communication will be greatly inhibited.
Other advantages and disadvantages have been identified from numerous studies of e-Education in diverse fields (see Table 2-2).
TAXONOMY OF COGNITIVE DOMAIN
Beginning in 1984, a group of educators undertook the task of classifying educational goal and objectives. The intent was to develop a classification system for three domains: the cognitive, the affective, and psychomotor. Work on the cognitive domain was completed in 1956 by Bloom and commonly referred to as Bloom’s Taxonomy of the Cognitive Domain that is now most widely used ways of categorizing level of abstraction of questions that commonly occur in educational settings. The major idea of the taxonomy is that educator may organize knowledge following a complexity hierarchy from lesser to more complex concepts. The taxonomy is demonstrated in table 2-4 with sample verbs for each level.
INTELLIGENT TUTORING SYSTEM
This section briefly examines the state of the art on the main components of traditional intelligent tutoring system (ITS) and gives some summary analysis for its limitations as well as its implication to MAS based ITS.
Computer has been used in education for over 20 years, Computer-based training (CBT) and computer-aided instruction (CAI) were the first such systems deployed as an attempt to teach using computers. While both CBT and CAI may be somewhat effective in helping learners, they do not provide the same kind of individualized attention that a learner would receive from a human tutor (Bloom 1984). For a computer based educational system to provide such attention, it must reason about the domain and the learner. This has prompted research in the field of intelligent tutoring systems (ITSs). Especially, the Carbonell’s (1970) proposal SCHOLAR system created an historical framework for ITSs that are computer-based instructional systems with models of instructional content that specify what to teach, and teaching strategies that specify how to teach (Wenger 1987, Ohlsson 1987).
The concept known as ITS or ICAI (Intelligent Computer-Aided Instruction) has many roots in Education, Psychology, and Artificial Intelligence (see figure 2-1). Nowadays, prototype and operational ITS provide practice-based instruction to support corporate training, college education, and military training.
The goal of an ITS is to provide the benefits of one-to-one instruction automatically and cost effectively. Unlike other computer-based training technologies, ITSs assess each learner’s knowledge, skills, and expertise. Based on the learner model, ITS can tailor instructional strategies, in terms of both the content and style, and provide explanation, hints, examples, demonstration, and practice problems as needed. By contrast, to CBT, ITSs offer more considerable flexibility in presentation of material and a greater intelligent mechanism to adapt to learner individual needs. These systems achieve their intelligence by making inferences about a learner’s mastery of topics or tasks in order to dynamically adapt the content or style of instruction. Content models (or knowledge bases, or expert systems, or simulations) give ITSs depth so that learners can “learn by doing” in realistic and meaningful contexts. Models allow content to be generated on the fly. ITSs allow « mixed-initiative » tutorial interactions, where learners can ask questions and have more control over their learning. Instructional models allow the computer tutor to more closely approach the benefits of individualized instruction by a competent pedagogue.
However, an ITS will typically constrain the learner to learn by a predetermined method or strategy (Rid 1989 & Kin 1997). ITS uses a model of the learner’s knowledge (learner model) so that the learner is presented with new information only when he/she requires it. This is carried out in order to reinforce a point, to progress in the learning and/or to identify misconceptions and wrong-rules (Sle 1982). Such systems have been criticized for constraining the learner to solving a problem in a particular way (Rid 1989). In most complex problem domains, there can be many methods to achieve a correct solution. Some people may find one particular method that suits their way of thinking better than others, it has been argued that learners should be able to experiment with their own ideas and find methods that naturally suit them.
Literature shows that a number of ITSs implement the cognitive tutoring strategy (Koedinger 2001; Anderson 1995). While most of today’s ITSs may appear to be monolithic systems, for the purposes of conceptualization and design, it is often easier to think about them as consisting of several interdependent components. Previous research by Wanger (1984), Woolf (1994) and Oliveira (1994) has identified four major components: (1) the expert module containing the domain knowledge, (2) the learner module, which accumulates information about the learner’s knowledge, misconceptions and behavior, (3) the pedagogic module, which includes the pedagogical expertise, and (4) the interface module (Figure 2-2).
MOBILE AGENT
Today’s most widespread paradigm for distributed computing follows the client-server paradigm. In this paradigm, the server is defined as a computational entity that provides some services. The client requests the execution of these services by interacting with the server. After the service is executed, the result is delivered back to the client. The server therefore provides the knowledge of how to handle the request as well as the necessary resources. A limitation of the client-server model is that the client is limited to the operations provided at the server. Therefore, if a client needs a service that a particular server does not provide, the client must find a server that can satisfy the request by sending out messages to all servers. This clearly is an inefficient use of network bandwidth. As a valuable alternative to the traditional programming model, the mobile agent paradigm involves the mobility of an entire computational entity, along with its code, state and probably the potential resources (e.g., ontology schemas) from host to host on a network. Dag (1994) implemented a computational metaphor that is analogous to how most people conduct business in their daily lives: visit a place, use a service, and then move on. Generally, the itinerary map whereby to travel through different network nodes can either be predefined or determined by the mobile agent on the fly, based on the its current state or its computing logic.
As compared with the traditional client-server model, the mobile agent paradigm has several advantages shown as below (Lange 1999, Wong 1999, and Chess 1998):
y Communication latency and bandwidth: If the communication between two interacting entities involves a considerable amount of data, it may be beneficial to move one of them close to the server instead of moving the data between them. The locality of the two entities allows them to decrease the latency and save bandwidth in the communication. Based on the locality information the agent may decide to move to the server location instead of invoking the server functions remotely. Since the interaction is carried out locally, it is independent of the network traffic.
Table of contents :
1HCHAPTER 1 INTRODUCTION
2H1.1 BACKGROUND &MOTIVATION
3H1.2 KEY ISSUES CONSIDERED IN THIS DISSERTATION
4H1.3 OBJECTIVES
5H1.4 THESIS ORGANIZATION
6HCHAPTER 2 STATE OF THE ART: E-EDUCATION, PEDAGOGIC THEORIES & MAS
7H2.1 INTRODUCTION
8H2.2 LITERATURE ON E-EDUCATION
9H2.2.1 Definition
1 0H2.2.2 Evolution of e-Education
1 1H2.2.3 Advantages & Disadvantages
1 2H2.2.4 Trend of e-Education
1 3H2.3 COGNITIVE THEORY IN EDUCATION
1 4H2.3.1 Cognitive Process
1 5H2.3.2 Taxonomy of Cognitive Domain
1 6H2.3.3 Learning Style
1 7H2.3.4 Constructivism
1 8H2.4 INTELLIGENT TUTORING SYSTEM
1 9H2.5 MULTI-AGENT SYSTEM
2 0H2.5.1 Background
2 1H2.5.2 Agent
2 2H2.5.3 Multi-agent System
2 3H2.5.4 Mobile Agent
2 4H2.5.5 FIPA Standard
2 5H2.6 SUMMARY
2 6HCHAPTER 3 ARCHITECTURE OF MAGE
2 7H3.1 INTRODUCTION TO E-EDUCATION REFERENCE MODEL
2 8H3.1.1 A recommended e-Education reference model
2 9H3.1.2 Leaning Technology Systems Architecture (LTSA) of IEEE LTSC
3 0H3.1.3 LTSA Overview
3 1H3.1.4 Stakeholder perspectives
3 2H3.2 FRAMEWORK OF MULTI-AGENT E-EDUCATION SYSTEM (MAGE) 1 46H51
3 3H3.2.1 classifications of agents in MAGE
3 4H3.3 LEARNING SCENARIOS
3 5H3.3.1 Scenario 1—Agent enabled intelligent tutoring system (AITS)
3 6H3.3.2 Learning scenario 2—Teacher intervened learning
3 7H3.4 SUMMARY
3 8HCHAPTER 4 MAS BASED COURSE & EEO AUTHORING
3 9H4.1 INTRODUCTION
4 0H4.2 DESIGN PRINCIPLAND CONCEPTMODEL
4 1H4.3 LEARNINGOBJECT DESIGN
4 2H4.3.1 DEFINITION LEARNING OBJECT
4 3H4.3.2 STRUCTRUE MODEL OF LO
4 4H4.3.3 EXTENSION OF LO METADATA TO ENHANCE ADAPTIVIEY
4 5H4.3.4 PACKAGE MODEL
4 6H4.4 ARCHITECTURE STRUCTURE
4 7H4.5 COURSEAUTHORINGSCENARIOS
4 8H4.5.1 SERCHING LEARNING OBJECTS
4 9H4.5.2 SUBSCRIPTION
5 0H4.5.3 NEGOCIATON WITH LO CREATORS
5 1H4.6 SUMMARY
5 2HCHAPTER 5 MAS BASED ADAPTIVE & ACTIVE LEARNING FRAMEWORK 1 66H77
5 3H5.1 INTRODUCTION
5 4H5.2 DOMAIN MODELING
5 5H5.3 ADAPTIVE INDIVIDUAL LEARNING
5 6H5.3.1 agent Architecute
5 7H5.3.2 AUTOMATICE LEARNING PATH GENERATION
5 8H5.4 ADAPTIVE COLLECTIVE LEARNING
5 9H5.4.1 INTRODUCTION
6 0H5.4.2 peer help MODELING
6 1H5.5 LEARNER GROUPFORMINGMODELING
6 2H5.5.1 INTRODUCTION
6 3H5.6 SUMMARY
6 4HCHAPTER 6 AN INNOVATIVE E- ASSESSMENT APPROACH: MOBILE AGENT BASED PARADIGM
6 5H6.1 INTRODUCTION
6 6H6.2 OVERALL FUNCTION STRUCTRUE
6 7H6.3 PROTOTYPE DESIGN OF GENETIC ALGORITHM BASED MAS TEST
GENERATION SYSTEM (GAMASTG)
6 8H6.3.1 introduction
6 9H6.3.2 genetic algorithm
7 0H6.3.3 test ontology design
7 1H6.3.4 design of ga
7 2H6.3.5 Architecture
7 3H6.3.6 state Chart
7 4H6.3.7 interactive model
7 5H6.4 DESIGN OF TEST DELIVERY
7 6H6.5 DESIGN OF EVALUATION &RESULT PUBLISHING
7 7H6.6 SUMMARY
7 8HCHAPTER 7 MAS IMPLEMENTATION & SIMULATION BASED ON JADE FRAMEWORK
7 9H7.1 INTRODUCTION
8 0H7.2 JADE
8 1H7.2.1 introduction
8 2H7.2.2 jade architecure
8 3H7.3 IMPLEMENTATION OF GAMASGT
8 4H7.4 IMPLEMENTATION OF TEST ONTOLOGY WHITH PROTEGE
8 5H7.4.1 7.4.1 design of agent behavior model
8 6H7.4.2 7.4.2 agent implementation
8 7H7.4.3 7.4.2.1 generic agent internal architecture
8 8H7.4.4 7.4.2.2 Implementation of Teacher agent
8 9H7.4.5 7.4.2.3 implementation of test generaration service agent (TGSAgent) 2 03H138
9 0H7.4.6 7.4.2.4 implementation of GActrlagent
9 1H7.4.7 7.4.2.5 TPagent
9 2H7.4.8 7.4.3 platform implementation
9 3H7.4.9 7.4.3.1 simulation
9 4H7.5 IMPLEMENTATION OF LEARNER MODEL AGENT
9 5H7.5.1 Protocoal implementation
9 6H7.5.2 Scenario
9 7H7.6 IMPLEMENTATON OF PEER HELP SYSTEM
9 8H7.7 SUMMARY
9 9HCHAPTER 8 CONCLUSIONS AND PERSPECTIVES
1 00H8.1 CONCLUSIONS
1 01H8.2 PERSPECTIVES
1 02HACKNOLEDGEMENT
1 03HREFERENCES