Research On Weighted Community Network Division Algorithm

Research on complex networks has gradually emerged since the end of the twentieth century,and it is rapidly penetrating into research in various fields,causing widespread concern among researchers in mathematics,physics,computers,and sociology.The problem of community structure detection is an important branch of complex network research,and it has gradually become a hot spot in complex network research.The research of unweighted community networks in complex networks can no longer meet the current research needs for network analysis,and the weights of edges in the weighted community network represent the relationship of the strength of connections between nodes.The study of the weighted community network partition algorithm helps people Understanding of real networks and discovering hidden information in weighted community networks.The main research con tents of this article are as follows:(1)In order to extract the important nodes in the weighted community network,the weight of the network edge is integrated into the calculation of the efficiency value of the structural hole theory,and an important node extraction algorithm based on the structural hole theory is proposed.The algorithm first combines the k-shell method with the improved structure hole efficiency value to calculate the importance of the nodes,and then sorts the nodes in descending order according to the importance,so as to achieve the purpose of extracting important nodes.Finally,the experiments conducted on the Zachary's Karate club dataset show that the effectiveness of the algorithm in extracting important nodes is significantly improved compared with the centrality and near-centrality algorithm.(2)Aiming at the problems of the current community network divisioning algorithm based on similarity,which do not consider common neighbor nodes,ignore the weight of edges,and have high time complexity,this paper proposes a weighted community network division algorithm based on common neighbor nodes and proximity.The algorithm starts with important nodes and combines the weighted local modularity,and then divides the weighted community network through clustering.Experiments on the simulation data set and the real data set show that for the simulation data set,the time used by the algorithm is 2.86%lower than the CRMA algorithm;for the Zachary's Karate club data set,the community modularity after the algorithm is divided The CRMA algorithm is improved by 4.51%.(3)In order to improve the accuracy of the weighted community network division algorithm,a weighted community network division algorithm based on genetic algorithm is proposed.The algorithm combines the merger operator and the split operator into the genetic operation,uses the weighted modularity function as the fitness function,and uses the optimization operator to find the optimal solution,thereby improving the accuracy of the weighted community network division.Finally,the experiment was conducted on the Dolphin network data set.In terms of modularity,the algorithm is 4.15%higher than the weighted FN algorithm and 8.85%higher than the WGN algorithm.