A FUZZY LOGIC-BASED INTELLIGENT TUTORING SYSTEM (ITS)

Most current Intelligent Tutoring Systems (ITS) use rule-based inference engines in order to select a prototype remediation for students1. Even Though this method has shown a great accomplishment in the area of artificial intelligence, many system experts are not satisfied with the rigidity of rule-based methods. Recently, research on possibilities for assessment using computer-based apprenticeship environments has been conducted by a team at the Learning and Development Research Center, University of Pittsburgh, suggests the possibility of using fuzzy variables instead of ordinary rule-based leveling methods. This paper describes an innovative technique for using a ''fuzzy relations'' approach to knowledge inference. This hybrid approach to the intelligent module of the ITS should prove to be a tremendous asset for timely individualized learning. Fuzzy relations not only provide speed in decision support over rule based inference engines, but a more experienced control of learning sequence, material and remediation.