Research On Overlapping Community Detection Algorithm Based On Node Distance Of Line Graph

Nowadays,with the rapid development of the era of big data,there is an increasing demand for data processing.Most systems in the real world can be abstracted into complex networks.A particularly important structure in a complex network is a community,which is a collection of individuals with common characteristics,which is widespread in various complex networks.For example,users with common interests in social networks can often become the same community.Excavating the community structure in the network accurately and efficiently will not only help us better understand the topology of the network,but also have a more comprehensive understanding of the internal laws of complex systems.However,some problems in the process of community mining can lead to poor quality of community detection,such as difficulty in accurately selecting community centers and overlapping characteristics between communities.Therefore,the key problem to be solved in the research of community detection is how to accurately and efficiently determine the node of community center and design the algorithm of the topology structure information of fusion network.The line graph is a model that can be widely used in edge community detection.The obvious advantage of edge community detection is that it can directly obtain the overlapping community structure in the network.Based on the above theory,this paper proposes an overlapping community detection algorithm based on node distance of line graph(OCDNG).This algorithm takes into account the topological structure of the links in the graph,and calculates the node distance between the line graph nodes by using the line graph to convert the edges of the graph into the nodes of the line graph.After obtaining the node distance matrix of the line graph,this paper we proposed a new ? measure to improve the fast search density peak clustering algorithm,and identify the node community of the line graph.Finally,the node community of the line graph is transformed into a graph,and the overlapping node community is obtained.Under the real social networks and biological networks,this paper proposed the algorithm with several classic community detection algorithm experimental comparison and analysis,the results show that this algorithm in the normalized mutual information to measure the NMI and extended modularity EQ indicators performance is good,not only improve the accuracy of the community division,also ensure the stability of the algorithm.