Abstract: Micro-blog gathers a large number of users, and community structure is a typical feature of Micro-blog network. In order to explore the influence of community structure on the opinion topic propagation of Micro-blog, users' psychology was analyzed for selecting their followees, and a model for constructing Micro-blog network was proposed which could modify community structure of network with a community factor. Experiments testify the effect of the community factor on micro-blog network structure, and use a SIR model to analyze the influence of different topic source nodes and community factors on the topic propagation of simulated networks. The result shows that community structure has a negative effect on the topic propagation compared with the effect of celebrities in the Micro-blog. An example data from twitter testifies some conclusions.

Key words: micro-blog, topic propagation, community structure, network modeling