Recommendation system for E-learners: model-based collaborative filtering approach

The e-learning recommendation systems help learners in making choices based on their personal preferences, even if they don’t have enough experience to compare different alternatives. Implementing a recommendation system in an online learning environment, that provides learners with several choices based on their personal experience. In this article, an intelligent adaptive e-learning system has been proposed, combining machine learning and reinforcement learning. This system aims to offer an individualized learning path to a specific learner to recommend an adaptive learning path, by applying the Collaborative Filtering algorithm based on SVD and data mining. The proposed system is broken down into three main modules, the data pre-processing module, the model creation module based on learner preferences, and the learning path recommendation module. The interactions between these three modules allow a personalized adaptation according to the preferences of the learner. Furthermore, we conducted some experiments to evaluate the effectiveness of our contribution. The proposed strategy demonstrated that by taking into account the learner’s personal preferences, the student’s learning experience can be improved and strengthened during the learning activity.