A proposed student model algorithm for an intelligent tutoring system

Tutoring is a linguistic exchange whose goal is to clarify a body of knowledge to which the student has already been exposed. Tutoring also involves directly a dialog so that the responses remain appropriate even when facing errors. This paper presents a proposed algorithm for the student model which can be considered as one of the main modules existing in an intelligent tutoring system. The proposed algorithm takes care of transferring knowledge to the student, pointing out the student errors, and understanding the student belief. This can be done by keeping track of the student status. The student model will also monitor the student behaviour. The monitoring process can be done by analyzing the student answers, comparing them with the correct system generated answers. This comparison will predict the student level of understanding or to recognize the student particular learning style. The student model is supported by a powerful structured knowledge base (KB) which contains a large number of facts and rules of the selected domain. The KB contains diagnostic hypotheses which are able to explain the student behaviour. The student model using that KB can perform logical inference operations to detect the student exploration direction, the student weak points, the student belief, and others.