Towards an Optimized Design of Individualized Learning Paths: an Approach Based on Ontology and Multi-agents System

In this paper we present an intelligent architecture, oriented goals, to create individualized learning paths. The adaptation of learning paths to learner profiles is an area of research growing. More research in this field has shown that taking into account the preferences and learning styles of learners improve the quality of the teaching/learning; thus, the collection of information characterizing learners as, for instance, preferences, learning styles, goals ... etc, and those characterizing learning resources (annotation of resources) are essential in order to make a matching between the query of learners and the profiles of hypermedia learning units. To recover their learning style, the learner is asked to take a test based on the model of Felder and Silverman. This test tells us about cognitive characteristics and affective behaviors and psychological which serve as relatively stable indicators of how learners perceive, interact and react with learning environments. Our contribution, therefore, consists of an adaptive approach based on semantic web, multi-agent systems and neural networks; thus, providing learners with personalized courses according to their profiles and their learning objectives.