A Student Model to Assess Self-Explanation while Learning from Examples

The SE-Coach is a tutoring system that supports students in applying the learning strategy known as self-explanation the process of clarifying to oneself the solution of an example. In this paper, we describe the student model used by the SE-Coach to assess the students’ self-explanations and to provide hints to improve them. The assessment is based on the student’s prior physics knowledge and on the student’s studying actions. We describe a version of the user model based on a Bayesian network, and a simplified version that is more efficient but handles only examples with no inferential gaps in the solution.