Toward a Hybrid Recommender System for E-learning Personnalization Based on Data Mining Techniques

Personalized courseware authoring based on recommender system, which is the process of automatic learning objects selecting and sequencing, is recognized as one of the most interesting research field in intelligent web-based education. Since the learner’s profile of each learner is different from one to another, we must fit learning to the different needs of learners. In fact from the knowledge of the learner’s profile, it is easier to recommend a suitable set of learning objects to enhance the learning process. In this paper we describe a new adaptive learning system-LearnFitII, which can automatically adapt to the dynamic preferences of learners. This system recognizes different patterns of learning style and learners’ habits through testing the psychological model of learners and mining their server logs. Firstly, the device proposed a personalized learning scenario to deal with the cold start problem by using the Felder and Silverman’s model. Next, it analyzes the habits and the preferences of the learners through mining the information about learners’ actions and interactions. Finally, the learning scenario is revisited and updated using hybrid recommender system based on K-Nearest Neighbors and association rule mining algorithms. The results of the system tested in real environments show that considering the learner’s preferences increases learning quality and satisfies the learner.