Social Analytics Framework to Boost Recommendation in Online Learning Communities

Online learning communities have become an important place serving informal learning due to the prevalence of online social networking services during the past few years. This paper proposes a social analytics framework aiming to boost recommendation service catering for the different learning demands of learners. Based on the traditional collaborative filtering approach, this study focuses on constructing topic-specific user credibility network by considering social relations and user behaviors. Both direct and indirect connections evidence from social analytics provide complementary information to construct user trust network. Regarding the topic-specific user credibility network, two features including influence and expertise are also computed to refine the credibility value between users. Furthermore, the performances of learners were further investigated in terms of longevity and centrality that could be referred when selecting suitable people for recommendation.