Measurement and analysis of online social networks

Abstract

Social network sites, like Facebook, Twitter, Renren and Sina Weibo, are now becoming increasingly popular on the Internet. For the past few years, numerous research have been made to investigate the topological structure and user behaviors of online social networks, which is quite important for the understanding of human social behaviors, the improvement of current Website systems and the design of online social networks' new applications. This paper provides an overview of online social networks' topology, user behaviors and network evolution. It also summarizes several common measuring methods and typical topological features; highlights user behavior characteristics and their impacts on network topology, and the network evolution. The conclusion can been drawn that as research progresses, the new characteristics of online social networks are gradually recognized and understood: users with a smaller number of correspondents tend to interact more with a subset of correspondents, while users with a very large number of correspondents actually spread their activity evenly across all of the correspondents; users' interactions decrease the clustering coefficient and loose the connections between neighbors; edge creation is influenced by both preferential attachment and proximity bias; small communities tend to merge with large ones which tend to split into two comparable size communities.