Distributed Data Distribution Mechanism in Social Network Based on Fuzzy Clustering

With the development of Internet, especially the success of social media like Facebook, Renren and Douban, social network has become an important part of people’s life. For social applications, one of the key problems to be solved is how to distribute data accurately to different users and groups in high speed. In this paper, we introduce fuzzy clustering into social network analysis. Users with similar interests are clustered into the same network according to fuzzy similarities. Our goal is to study how such clustering can enhance data distribution to the largest extent. We take 3,000 user’s data from Douban.com, fetch fuzzy social relationship, and simulate data transmission with a theme-based pub/sub mechanism. Experiments show that network clustering based on fuzzy clustering can improve data distribution effectively, while remain robust in highly dynamic environment.