Social Recommendation Algorithm Based on Trust and Distrust

Recommendation system can alleviate the problem of information overload effectively and enable users to obtain target information quickly. However, data sparsity is one of the important factors affecting the accuracy of recommendation system. The matrix decomposition algorithm based on trust makes use of the rating information and trust relationship between users to make recommendations. However, most matrix decomposition algorithms based on trust only consider the explicit trust relationship between users, and rarely consider the distrust relationship and implicit social relationship between users. Therefore, this paper proposes a new recommendation algorithm, which makes use of the rating information between users, explicit social relations and implicit social relations obtained through the rating information to recommend, which can alleviate the problem of data sparsity to a large extent. Finally, experiments with real data set Epinions show that the proposed algorithm can further improve the accuracy of the recommendation system.