Friend Recommendation Method using Physical and Social Context

Social network sites have attracted millions of users with the social revolution in Web 2.0. In the social network sites, a user can register other users as friends and enjoy communication. However, users of social network sites may easily get overwhelmed by the excessive volume of friend information. Recently context-aware mobile devices have been thoroughly integrated into all walks of life. The context-aware systems provide the user with adaptive recommendations from enormous information. Therefore, the essential factor of social computing is to recommend truly valuable friends using context. We propose a friend recommendation method using physical and social context. The main idea of the proposed method is consisted of the following three stages; (1) computing the friendship score using physical context; (2) computing the friendship score using social context; (3) combining all of the friendship scores and recommending friends by the scoring values.