Date of Award

Summer 2001

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Industrial Engineering - (Ph.D.)

Department

Industrial and Manufacturing Engineering

First Advisor

Golgen Bengu

Second Advisor

Jack Gelfand

Third Advisor

One-Jang Jeng

Fourth Advisor

Barbara B. Kebbekus

Fifth Advisor

Arijit K. Sengupta

Abstract

In this study, two components are developed for the Web-based adaptive learning: an online Intelligent Tutoring Tool (ITT) and an Adaptive Lecture Guidance (ALG). The ITT provides students timely problem-solving help in a dynamic Web environment. The ALG prevents students from being disoriented when a new domain is presented using Web technology. A prototype, Distributed Intelligent Learning System (DISTILS), has been implemented in a general chemistry laboratory domain.

In DISTILS, students interact with the ITT through a Web browser. When a student selects a problem, the problem is formatted and displayed in the user interface for the student to solve. On the other side, the ITT begins to solve the problem simultaneously. The student can then request help from the ITT through the interface. The ITT interacts with the student, verifying those solution activities in an ascending order of the student knowledge status. In DISTILS, a Web page is associated with a HTML Learning Model (HLM) to describe its knowledge content. The ALG extracts the HLM, collects the status of students' knowledge in HLM, and presents a knowledge map illustrating where the student is, how much proficiency he/she already has and where he/she is encouraged to explore. In this way, the ALG helps students to navigate the Web-based course material, protecting them from being disoriented and giving them guidance in need.

Both the ITT and ALG components are developed under a generic Common Object Request Broker Architecture (CORBA)-driven framework. Under this framework, knowledge objects model domain expertise, a student modeler assesses student's knowledge progress, an instruction engine includes two tutoring components, such as the ITT and the ALG, and the CORBA-compatible middleware serves as the communication infrastructure. The advantage of such a framework is that it promotes the development of modular and reusable intelligent educational objects. In DISTILS, a collection of knowledge objects were developed under CORBA to model general chemistry laboratory domain expertise. It was shown that these objects can be easily assembled in a plug-and-play manner to produce several exercises for different laboratory experiments. Given the platform independence of CORBA, tutoring objects developed under such a framework have the potential to be easily reused in different applications.

Preliminary results showed that DISTILS effectively enhanced learning in Web environment. Three high school students and twenty-two NJIT students participated in the evaluation of DISTILS. In the final quiz of seven questions, the average correct answers of the students who studied in a Web environment with DISTILS (DISTILS Group) was 5.3, and the average correct answers of those who studied in the same Web environment without DISTILS (NoDISTILS Group) was 2.75. A t-test conducted on this small sample showed that the DISTILS group students significantly scored better than the NoDISTILS group students.

Share

COinS