Managing inference process in student modelling for intelligent tutoring systems

We present an approach to managing the side effect of information updates in the student model (SM). The approach is based on fuzzy logic and fuzzy reasoning on the knowledge structure in the SM. The SM includes three parts: a cognitive model, a behavioural model and an inference engine. The cognitive model is an overlay on curriculum knowledge structures, the behaviour model contains affective and conative values of the student and the inference engine aims to manage updates occurring in the SM during the learning process. As updates imply the evolution of the knowledge structure of the student the propagation module evaluates the possible impacts of the updated information on other related information. This is done by activating a fuzzy reasoning on knowledge networks in the curriculum and deducing new information in the SM. The control module handles conflict situations by considering factors such as the information source and the confidence degree