JavaTM Intelligent Tutoring System Model and Architecture

Accessibility of computers and computer resources are increasing in our society at a staggering rate. Computer technology is changing more rapidly now than at any other time in history and the price of computers are continually decreasing inversely proportional to the power they deliver. Nearly 50% of households in Canada and the United States have computers [15]. Internet connections and capabilities are growing at an amazing rate due to the number of people who want to be connected to the world of information [15]. Internet Service Providers are upgrading their infrastructure to support real-time video and audio to their clients. Personal Digital Assistants such as cellular telephones and palm-pilots are Internet-ready and becoming commonplace in our society. In spite of the advances in computer technology and accessibility, educators have been relatively slow in seizing technology to enhance student learning. There are significant problems in the context of personalized student instruction in current educational systems that can be remedied through the use of appropriate technologies. Online teaching tools such as WebCT and Blackboard are becoming extremely popular for distance and in-class education. In fact, entire universities have implemented online teaching tools as the central mechanism for delivering all of their courses [16]. The strength of these tools is their ability to provide the teacher and student with a great deal of versatility within the learning environment [16]. Unfortunately, they do not provide any means by which a student may receive ongoing personalized instruction. Teaching students on a one-on-one basis significantly influences the degree of knowledge and skill retained by the student; Bloom [3] showed that an individual human tutor can improve student learning by two standard deviations over classroom instruction. In other words, the average individually tutored student performs better than 98 percent of students in a classroom instructional environment [10]. This raises the following crisis in the educational community. In order for students to reach their potential they need individual tutoring. However, due to numerous limitations such as access to online teaching tools, financial considerations, and sheer logistics each student cannot be granted access to a personalized human tutor for a consistent duration of time. After all, traditionally there is only one teacher in a classroom of students. So, what can be done to solve this problem? One solution lies in the implementation of Intelligent Tutoring Systems. __________________________________ Copyright © 2002, American Association for Artificial Intelligence (www.aaai.org). All rights reserved. Introduction A technological counterpart to a human tutor called an Intelligent Tutoring System (ITS) may be an answer to the current crisis in education. ITSs are computer based systems designed using cognitive science and Artificial Intelligence (AI) techniques to provide students with individualized instruction in a similar way to human tutors. The “Java Intelligent Tutoring System” (JITS) model focuses on designing an accelerated learning system using Intelligent Tutoring Systems (ITS) coupled with advanced artificial intelligence components. The framework of JITS will be sufficiently flexible to provide minimal configuration to enable other subject disciplines to be plugged-in (e.g., calculus, geometry, biology, etc.). An additional design aspect of JITS is its augmentation to popular web-based instructional tools such as WebCT and Blackboard. The proposed model consists of two distinct components. First, an overview of the characteristics of existing Intelligent Tutoring Systems is described. Second, the design strategies of the Java ITS is presented. The JITS draws from the research advancements of ITS, cognitive science, and AI. This project is not yet complete. However, it is hypothesized that the Java Intelligent Tutoring System will provide an interactively-rich learning environment for students that will result in increased achievement. Based on the success of similar Intelligent Tutoring Systems, it is also hypothesized that these students will be able to learn the course material more quickly than students in traditional classroom environments. The purpose is to demonstrate that learning may be accelerated and cognitive development deepened using this tutor. At the core of an ITS is an AI module. The AI module is responsible for many tasks including capturing the student’s knowledge state, delivering an appropriate lesson, assessing and evaluating student performance, and providing valuable feedback to the student. Thus, an ITS is explicitly designed to assist in resolving the crisis in education regarding personalized student education by providing a means through which students may improve their academic performance. The purpose of this study is to design and construct an ITS using advanced AI components for students and to perform a quantitative study to determine ITS’ effectiveness. From: AAAI Technical Report SS-03-04. Compilation copyright © 2003, AAAI (www.aaai.org). All rights reserved.