A web mining based approach for automatic student model discovery

Learner models represent a basic knowledge asset that can be used to ensure personalization within e-learning systems. These models can be built only based on learners activities, tracked and gathered on the web server side. In this paper, we propose to outline the general principles of an entirely automated web mining based approach for modeling learners in learning management systems. So, we consider a learner model with three components: the learner´s profile, the learner´s knowledge, and the learner´s educational preferences. These learner´s model components are inferred automatically from usage data, based on web mining techniques. Then, a hierarchical multi-level model based collaborative filtering approach is applied for modeling learners into groups.