Thi Dieu Anh Nguyen, Thanh Nguyen Vu and Tuan Dinh Le
Abstract. Currently, most of the recommendation systems use user’s feedbacks to suggest an item for the user. However, the current recommendation system used only explicit rating information, which not fully evaluating the factors that directly affect the feeling of users in rating. To achieve more accurate results, this paper proposes a solution to add the implicit effect of items rating to the recommendation system based on the TrustSVD model and matrix factorization (MF) techniques. The experimental results showed our proposed solution achieve better than 18% the matrix factorization method and 15% the Multi-Relational Matrix Factorization method.
Keywords: Recommendation system, Collaborative filtering, Implicit effect, Matrix Factorizations, Matrix user, Trust-based recommender
https://link.springer.com/chapter/10.1007/978-3-030-35653-8_47