Effective Resource Recommendations for E-learning: A Collaborative Filtering Framework Based on Experience and Trust

A personalized e-learning recommender system can help learners in finding learning resources that suits their needs. Collaborative filtering (CF) generates recommendations to learners by leveraging the preferences of group of similar learners. In this paper, we put forward that in addition to the traditional similarity in recommending the E- learning resources (like book, subjects, teachers), other factors such as experience and trust have an important role in generating effective recommendations. So, we considered both experience and trustworthiness of learners to develop a collaborative filtering framework, called CF-EXP-TR scheme, for e-learning recommendations. A two level filtering methodology is proposed which shows that recommendation of learning resources is taken by those learners who are more experienced as well as high trustworthy. The experimental results show that incorporation of experience and trust concepts into collaborative framework, CF-EXP-TR indeed improves the recommendation accuracy and establishes that our scheme is better than traditional Pearson based collaborative filtering, CF-PR.