Empirical Analysis and Evolution Modeling of Network Topological Structure in Microblog

Abstract

The wide use of Microblog leads to an instant online community, and the research on the Microblog networks topological structure is meaningful for understanding the information dissemination mechanism. We studied the distributions and correlation of the users' followers, friends, and bidirectional friend numbers and the correlation among them. In order to study the topological structure features of Microblog, we collected data from Sina Weibo and made a real bidirectional connection network. Using complex network theory, we analyze the statistical properties of this network, demonstrate that it processes small world and scale-free features. Moreover, we analyze some topological structure metrics, such as degree distributions, node degree correlation, and clustering coefficient distributions. Through inspecting the statistical properties, we find that it is disassortative and has hierarchy structure. In addition, we find that the users' age distribution can be divided into two sections and that there will emerge a large degree node in various stages of network evolution, but user average degree with user age has a gradual upward trend. We propose a fitness-based model with node accelerated growth, and the simulation results show that our model can be better consistent with the real network.