1 Introduction
In the face of a huge number of web users and information explosion, recommender systems are of vital importance which can alleviate information overload and provide users with more efficient and high-quality services. An effective recommender system can benefit both users by acquiring their preferred contents (e.g., movies, music, merchandise) from a large amount of information, and service providers by reducing promotional costs. As a result, recommender systems have attracted widespread interests in recent years. Meanwhile, exploiting social relations to improve the performance of recommendation has also become increasingly popular with the growth of social media [1], [2], [3]. In social networks, there is a flow of information among connected friends. A user’s preference is similar to or influenced by the people around him/her, which has been proved by social correlation theories [4], [5].