Study On Information Spreading Over Heterogeneous Network

With the development of Internet technology for the past few years, social networks play an important role in people's lives, which have become a part of our livelihoods and play an underestimable impact on information acquisition,people's thinking and living. Large amounts of information are generated in the network and propagated through huge social relationship chain every day, some of which, such as international and social news will spread quickly and exist for a long time, while other marginalized information only propagates in a small range and disappears soon. Moreover, it has become a public platform which provides service of sharing and releasing information. Users can update their status anytime and anywhere on the dating sites and interact with friends. Hence, it has become a main approach to obtain information, show oneself and act as marketing windows. However, it also has some drawbacks, for instance, it can be a center of rumor spreading. Therefore, it is of great significance to study the information propagation and control information spreading in social networks.Most of the current study about information spreading refer to the SIR model based on the mean-field hypothesis,which assume that all nodes have the same numbers of neighboring nodes, that is, the degree of each node is approximated by the average degree. However, the networks in the real world are mostly heterogeneous, which have different degree distribution, for example, the degree of users on facebook has a power-law distribution. As a consequence, it is great important to introduce the degree distribution of node for the study of information spreadng.This thesis explores the spreading mechanisms of information spreading over social networks, and builds a SIR-based spreading model according to the process of information spreading. With the node heterogeneity of the real social networks taken into consideration, the dynamic evolution equations for information spreading process are obtained on the basis of the node degree distribution of social networks. The final size of the information spreading is evaluated and the basic reproduction number and the early stability of the model are derived from the dynamic evolution equations. The effects of information spreading mechanisms and network models are investigated by numerical simulations and the results of theoretical analysis are validated.