Intelligent Tutoring Systems Based Upon Qualitative Model Evolutions

Barbara Y. White, John R. Frederiksen

One promising educational application of computers derives from their ability to dynamically simulate physical phenomena. Such systems permit students to explore, for instance, electrical circuit behavior or particle dynamics. In the past, these simulations have been based upon quantitative models. However, recent work in artificial intelligence has created techniques for basing such simulations on qualitative reasoning. Qualitative models not only simulate the phenomena of the domain, but also permit instructional systems to generate explanations of the behavior under study. Sequences of such models, that attempt to capture the progression from novice to expert reasoning, permit instructional systems to select problems and generate explanations that increase in complexity at an appropriate rate for each student. Since the acquisition of a qualitative understanding of the laws of physics and their implications is an important component of understanding physical phenomena, it is argued that systems based upon qualitative model progressions can play a valuable role in science education.

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.