Towards a Personalized E-Learning System

Towards a Personalized E-Learning System

Elvis Wai Chung Leung (City University of Hong Kong, Hong Kong) and Qing Li (City University of Hong Kong, Hong Kong)
Copyright: © 2009 |Pages: 11
DOI: 10.4018/978-1-60566-198-8.ch314
OnDemand PDF Download:
$37.50

Abstract

To cope with the increasing trend of learning demand and limited resources, most universities are taking advantage of Web-based technology for their distance education or e-learning (Montelpare & Williams, 2000). One of the reasons is due to the significant price drop of personal computers in recent decades; the Internet and multimedia have penetrated into most households. Moreover, most students prefer to learn from an interactive environment through a self-paced style. Under the Web-based learning model, students can learn anytime, anywhere because they are not required to go to school on schedule (Appelt, 1997). Meanwhile, universities also enjoy the economic benefit due to the large student base that can share the development cost of course materials and other operational expenses. Gradually, more and more universities follow this similar way to provide online education.
Chapter Preview
Top

Research Background

Over the past decade, some researchers have focused on the areas of agent-based, multimedia, and adaptive approaches for e-learning systems. Before we introduce our PEELS approach, some related research projects are discussed below.

For agent-based research work, Andes (Gertner & VanLehn, 2000), a collaborative project started in 1995 between the University of Pittsburgh and the U.S. Naval Academy, has taken a modular architecture and is implemented in Allegro Common Lisp and MS Visual C++. The main agents of Andes contain a problem author, a problem solver, an action interpreter, and a help desk. These agents function in the sequence of creating a problem definition, generating a problem-solution graph model, and referring to the student model to make decisions upon receiving appropriate feedback and assistance. LANCA (Frasson, Martin, Gouarderes, & Aimeur, 1998) adopts an intelligent agent’s approach to distance learning in a distributed environment by using a constructive approach instead of user profiles to assist students in difficulty by building a common and useful data bank. Unlike Andes and LANCA, DT Tutor (Murray & VanLehn, 2000) uses decision-theoretic methods for coached problem solving to select tutorial actions that are optimal given the tutor’s beliefs and objectives. It employs a model of learning to predict the possible outcomes of each action, weighs the utility of each outcome by the tutor’s belief that will occur, and select the action with the highest expected utility.

Key Terms in this Chapter

Course-Material Structure: As the course materials can be stored in XML (extensible markup language) format, which offers a tree-like or hierarchical structure, the prevailing relationships inside the course materials are of parent-to-child, which facilitates easier addition or removal of the course-material nodes or documents. Through techniques like SMIL (synchronized multimedia integration language), multimedia data not only can be played on the Internet, but also run in a synchronized manner.

MultiAgent System: It can be considered as an ensemble of agents acting and working autonomously, each representing an independent focus of control of the whole system. Moreover, it is a loose network of problem-solver entities, and the entities work together to find answers to problems that are beyond the individual capabilities or knowledge of each entity.

Dynamic Conceptual Network: It is a hierarchical tree for developing the interrelations among learning concepts. Each learning concept is stored as a concept node in the dynamic conceptual network. A concept node includes contents, tasks, and attributes. The contents are represented by text, graphs, audio, and/or video, and these contents are stored in a knowledge base.

User Profile: The profile contains all the data associated with the user or learner, for example, educational background, preferences, learning aims, competitive level, and so forth.

Knowledge Base: It is the primary repository for all course materials. Relevant information from the knowledge base will later be extracted out in providing personalized course materials for individual students.

User View: It is a subset of the external schema for querying the course-material knowledge base; the subset is usually defined based on the corresponding user profile for extracting relevant data and information.

Agent: It is an encapsulated computer program situated in some environment, and is capable of flexible, autonomous actions in that environment in order to meet its design objectives.

Complete Chapter List

Search this Book:
Reset