Building Learner Profile for Group Learning Recommender System from Learning Process

Abstract

We propose a learning process based approach to address the well-known implicit profiling method in group learning recommender system. Our approach makes use of likelihood control to integrate the learning step coefficient of each learner into ratings. We introduce a preference model comprising both user-user and user-item relationships in recommender systems, and present a motivating example of our work based on the model. We empirically evaluated the effectiveness of our approach, which shows that the proposed approach has a good recommendation performance in e-learning recommender system.