Abstract
Many computer-based training systems present instruction linearly, with exactly one path through the system that each student must follow. Students have little control over the pace, content branching, or flow of instruction. Student modeling within intelligent tutoring systems addresses these issues by interpreting student behaviors, representing the student's knowledge, and providing personalized instructional content. However, there is disagreement over the necessary content and structure of student models and their general utility. This paper discusses the development of an intelligent tutoring architecture with two instantiations that use sophisticated student models. Similar tutors with scaled-down student models will be used to evaluate the effectiveness of the differing student modeling approaches.
Get full access to this article
View all access options for this article.