Machine Learning Approaches for Inducing Student Models

This paper focuses on Machine Learning approaches for inducing student profiles, respectively based on Inductive Logic Programming (the INTHELEX system) and on methods using numeric algorithms (the Profile Extractor system), to be exploited in this environment.

Abstract

The main issue in e-learning is student modelling, i.e. the analysis of a student's behaviour and prediction of his/her future behaviour and learning performance. Indeed, it is difficult to monitor the students' learning behaviours. A solution is the exploitation of automatic tools for the generation and discovery of user profiles, to obtain a simple student model based on his/her learning performance and communication preferences, that in turn allows to create a personalized education environment. This paper focuses on Machine Learning approaches for inducing student profiles, respectively based on Inductive Logic Programming (the INTHELEX system) and on methods using numeric algorithms (the Profile Extractor system), to be exploited in this environment. Moreover, an experimental session has been carried out, comparing the effectiveness of these methods along with an evaluation of their efficiency in order to decide how to best exploit them in the induction of student profiles.