A Hybrid Recommendation System Based on Correlation and Co-Occurrence Within Social Learning Network

Social learning is one of the most prevalent disciplines in terms of e-learning. To handle learning resources within social environments, recommendation systems are gaining tremendous prominence based on a series of criteria such as the rate of learner interaction with the learning environment. On the basis of this, we highlight an overriding issue focusing on the influence of the rate of learner interaction on the calculated recommendations. In other words, to what extent considering the events carried out by the learners and the existing links between them will lead to more relevant and reliable recommendations. To emphasize this point and to support current recommendation systems, we are evaluating our recommendation approach that integrates a set of learner activities based on correlation and co-occurrence. We then compare the performance of the hybrid system to the following two recommendation systems: the recommendation system based uniquely on correlation and the recommendation system based solely on co-occurrence.