Sampling Online Social Networks for Analysis Purpose

Online Social Network is a novel communication pattern of great convenience and abundant interest, influencing people's life in extensive regions. A prominent problem in OSN is the access to data form a whole view of network topology from various social network sites. This paper describes the fundamental method to obtain a representative sample of online social network. According to the analysis of user interactions with social networking sites and the comparison of different traversal algorithms, this paper simulates the social network traversal based on the breadth-first algorithm. Then, we implement the algorithm on real OSN by sampling. A considerable data set is available by our sampling method. Finally, several social network analysis are experimented to characterize key properties of OSNs.