Research On Weibo-Oriented Varying-And-Selective Fitness Model

In recent years, complex networks science has rapidly developed. As a typical type of complex networks, Online Social Networks (OSNs) have become more and more popular. A large number of OSNs have appeared in our daily lives. The connections in the OSNs reflect the people's actual relationships in reality society, so more and more researchers have been interested in the structure and evolution of OSNs. It is of great theoretical value and practical significance to deeply understand the topology and evolution mechanism of OSNs, which can help to understand the propagation law of information on the Internet.Complex network theory is an important way to understand the structure and dynamics characteristics of complex system. It has been widely applied to the study of OSNs. In this thesis, the crawled social network data of Sina Weibo is analyzed first. It is found that the influence of many factors on the connection or disconnection between nodes is time varying, no matter the users'joining time, activity, "celebrity effect", or their similarity, the interactions between them. Additionally, the attraction of a node to different other nodes is different at the same time. Therefore, basing on fitness model, the variability and differentiation of fitness are combined in this thesis. Finally, Varying-and-Selective Fitness Model (VSFM) is proposed for modeling the evolution of networks of Weibo.VSFM reflects the mechanisms of growing and preferential attachment of the real social network in. It also reflects the phenomenon of "The Latecomers Surpass the Formers" for the node with higher fitness and more other similar nodes. In this model, preferential attachment, random adding or deleting edges, and two-hop relations are all considered during the network evolution, which reflect the evolution mechanism of the real network. The simulation results show that VSFM reflects characteristics of social network on Weibo well, with the power-law degree distribution, the high clustering coefficient, and the small world phenomenon. As a result, VSFM provides the basis for the research on information propagation, user relationship, and personalized recommendations on OSNs.