A Fuzzy Student Modeling with Two Intelligent Agents

Abstract

A new fuzzy student modeling method with two intelligent agents, a diagnosis agent and a learning agent, are suggested by this article for several aspects (problems) of student modeling in Intelligent Tutoring Systems. Those problems are still difficult tasks for current student models to deal with, such as dynamic interaction, ambiguity, incompleteness, and inconsistency. The diagnosis agent incorporates the classification tree concepts to identify the student's misconceptions and to do score assignment that includes the score assignment of partial correctness. The learning agent is used to learn the features of the student's inconsistent behaviors for the system to take effective strategies to prevent the happening of the student's inconsistent behaviors. This article also integrates fuzzy theories and Hasse diagrams for student modeling. The theories of fuzzy sets are developed to model imprecision of the real world. The Hasse diagram can be used to present the dynamic feature and to reason for overcoming the problem of incompleteness.