社交网络影响力大化传播模型与算法研究.pdf

Research on Propagation Model and Algorithm for Influence Maximization in Social Network Abstract With the emergence and rapid proliferation of social network applications in 21st century, such as Twitter, Facebook, Flickr and so on, people are acquiring information not just from newspapers, broadcasting and TV any more. Nowadays, social network is becoming a more and more important medium that conveys information. As the research of influence maximization in social network plays a very important role for guiding practical application, it has become a concerned problem in computer science field. We focus our attention on the models and algorithms of influence maximization to display our research as follows. First of all, we formulate the problem of influence maximization in social network in details. In order to know how the information propagates in social network, some important information spread models and corresponding definitions are analyzed, and some algorithms are summarized so as to build the theoretical foundation for our research on propagation model and algorithm of influence maximization in social network. Secondly, we point out the irrationality about the assumption of those basic models. As we know, different parent nodes have different influence on their child node. Based on this idea, we propose a new PRP model in which the correlation and authority of nodes are considered. The experimental results show that our proposed model is more effective than traditional Linear Threshold