| Information networks has brought enormous information resources to our production and life.However,with the deepening of research needs,it is an extremely challenging task to study and deal with complex information networks from the perspective of traditional thinking and analysis.Structural hole is an important theory reflecting the key of social network nodes,studying the characteristics of network information dissemination and diffusion,and community detection,and so on.Its novel research perspective plays an important role in the analysis of social networks influence,the detection of weak nodes of network equipment,the control of public opinion,the complementary advantages of cross-border cooperation between enterprises or individuals,and the enhancement of the competitiveness of enterprises in reverse applications.This paper starts from the research thought of graph theory and network topology structure,and based on attributes information such as SPIG the shortest path increment),NCC(the number of connected components)and VAR(the number of nodes variance),to study,and analyze the structural hole detection algorithm,then we designed and implemented two structure-holes detection algorithms for complex networks,which are SNV algorithm and SNV-BC optimization filtering algorithm based on intermediary centrality.First of all,after in-depth study and analysis of the existing structure hole detection algorithm based on the shortest path increment,according to its shortcomings,we supplement the definition of new attributes and redefine the network model,then proposed a new structure hole detection algorithm.Secondly,aiming at the problem of high time complexity caused by more attributes factors in algorithm SNV.a new optimal filtering solution is put forward based on the comprehensive analysis of algorithm principle and experimental data.Finally,the effectiveness and efficiency of the proposed algorithm are validated in different types of network data sets by using NDCG evaluation algorithm and SIR information difftusion model.Gephi,Networkx and other visualization technologies and tools are used to further visualize and explore the experimental results.The experimental results show that the proposed algorithm SNV-BC can give consideration to both efficiency and accuracy in the analysis and calculation of complex networks.At the same time,the algorithm itself also shows high scalability in weight,interaction and other aspects. |
