Core-based community evolution in mobile social networks

Hao Xu,Weidong Xiao,Daquan Tang,Jiuyang Tang,Zhenwen Wang

DOI: https://doi.org/10.1109/BigData.2013.6691782

2013-01-01

Abstract:Community evolution in social networks attracts a lot of attention in recent years. Existing methods always depict the relationship of two nodes using the temporary connection. However, these temporary connections cannot be fully recognized as the real relationships when the history connections among nodes are considered. Cumulative stable contacts are proposed to depict the correlation among nodes. The whole process is divided into timestamps. At each timestamp, the community cores will be detected due to the variation of nodes and links firstly. Then, all nodes will be divided into a few of communities due to the community cores. Meanwhile, communities can be tracked through the incremental computing, which can help to recognize the evolving of community structure. Empirical studies on real-world social networks demonstrate that our proposed method can effectively detect stable community in mobile social networks.