The Analysis And Application Of Key Nodes In Complex Social Networks

In recent years,more and more attention is attached to the research of complex social networks,especially about the analysis of key nodes in the network.As the scale expands and the number of dimensions increases unceasingly in complex social networks,The existing algorithm,such as the algorithms of centrality,K-shell,PageRank and the like,can't meet the requirements in terms of accuracy,efficiency and applicability.As a result,how to analyze the key nodes in complex social networks accurately and efficiently becomes the core problem to research in this thesis.Introduced from the perspective of the scale of complex social networks,this thesis selects the index possessing higher influence to mine the key nodes so as to accomplish the understanding,control,protection and prediction for complex social networks,which has important research significance and application value.In this thesis,the main work and innovation points are as follows:1.To address the problem that existing algorithms are adopted for the analysis of key nodes in complex social networks of different scales but sacrifice accuracy or efficiency,we put forward an feasible method to analyze the key nodes.On the one hand,aimed at the analysis of key nodes in small and medium-scale networks,Weighted Degree Centrality is proposed to enhance the accuracy.On the other hand,with respect to large-scale networks,in order to ensure the accuracy of mining key nodes,hierarchical approach ought to be applied to improve efficiency.2.To address the problem of analyzing key nodes in small and medium-scale networks,we propose an algorithm named Weighted Degree Centrality(WDC).We define the weighted-degree of nodes given the weights of incoming edges and the out-degree of its neighbor nodes for the target node synthetically.Combining with its individual influence,we quantificat the importance of the node,thus improving the accuracy of analyzing key nodes immensely in small and medium-scale networks.In addition,the stability of WDC is also verified and analyzed in detail,and we summarize the research significance of providing the high-quality nodes as a date set for large-scale networks by means of designing comparative experiments on data of different magnitude in the same data set in this thesis.3.To address the problem of analyzing key nodes in large-scale networks,the relationship strength is introduced,which is modeled according to static and dynamic factors,and another algorithm named NodeRank is proposed with that.This method analyzes large-scale networks hierarchically to reduce the scale of networks by means of selecting high-quality nodes first of all,and then we consider the weights of edges between nodes to compute the importance of these nodes based on PageRank.Finally,through simulation experiments,we prove that the combination of these two methods not only ensure the accuracy but also improve the effciency immensely for the analysis of key nodes in complex social networks in this thesis.