Online Social Influence And Personalized Recommendation: An Empirical Study
Recent studies suggest that online social relations influence users' both product choices and ratings, which provides empirical support for incorporating online social influence into the recommender models. To incorporate such influence, we consider the ratings from a user's social circle as a source of implicit feedback from the user. We then integrate such feedback into the representative recommendation model of collaborative filtering to improve personalized recommendations. The evaluation results show that our proposed approach performs better than the pure ratings-based methods in terms of accuracy and coverage.
Proceedings of the Annual Meeting of the Decision Sciences Institute (DSI), 2017.
© Decision Sciences Institute, 2017
He, X., Karami, A. (2017). Online social influence and personalized recommendation: An empirical study. Proceedings of the Annual Meeting of the Decision Sciences Institute (DSI), Washington, D.C. https://decisionsciences.org/wp-content/uploads/2017/11/p1291808.pdf