The amount of data available on the web has exploded exponentially in recent years and continues to grow. These data, which are heterogeneous and distributed, have a significant number of technical and technological challenges, whether for storage, exploitation or analysis.

The aim of this work is to develop an ontology-based hybrid recommender system for student orientation in higher education. In this thesis, we are interested in the management of heterogeneous data from several universities, the representation of knowledge related to students' trajectories and their environment, and the use of data mining techniques to recommend appropriate top N universities and fields of study according to students’ profile.