Research And Implementation Of Community Discovery Algorithms Based On Complex Networks

In past few years,due to the obvious development of Internet,a variety of network applications have sprung up and people have been living in a variety of complex network environment.Therefore,the social network analysis has become a important topic that scholars from all walks of life are generally concerned about.The relevant analysis of complex network data sets,so as to find the existence of the key rules and information in the network data,have become the main means and purpose of the study of complex networks.In complex networks,the person is equal to the node,and the link between the person and the person is equal to the side of the node.A large number of studies have found that the community structure generally exists in complex networks.Community structure represents a set of nodes with similar properties or play similar roles.Usually,the nodes in the community have a closer relationship.Through the study of community structure,we can research and explore the complex network's internal structure and attribute deeply,so as to find its implicit rules and predict its behavior.Therefore,the discovery and research of community structure has become one of the important research topics in the field of computer science.This paper will analyze and research on the complex network,community structure,community discovery algorithm,social network analysis and other topics,mainly including the contents of several aspects:(1)To solve the problem of low accuracy of existing community detection methods,a community detection algorithm based on central nodes is proposed here.Central nodes of communities are found through the degree of each node and the similarity of nodes.Each community is optimized using local modules.The community structure of the entire complex network is got by classifying isolated nodes and overlapping community nodes to their community as far as possible based on node attraction.(2)According to the community discovery algorithm based on central node,each node plays the same role in the complex network by default,and each node is in the same weight in the module calculation,which leads to the problem that different nodes can not reflect their own role and value in the network.In this paper,the basic idea of PageRank algorithm is introduced,and the concept of core sub group is proposed.In order to reflect the different roles of different nodes in the complex network,the paper puts forward the concept of the weighted local community contribution degree also.And on this basis,this paper proposes a community discovery algorithm based on core sub clusters.On the basis of finding a community center node,the core group of the community is found through the node degree and so on.Then the weighted local module is used to optimize each community,and community optimization is carried out on the special node of the community.So as to obtain the community division of the whole network.(3)In this paper,the two algorithms have certain advantages in the application of the complex network.Through community partition experiments on complex network data sets,and the accuracy and running time of the algorithm are compared with the related community finding algorithm,it can be seen that the algorithm in this paper has a certain degree of superiority.