Parallel Algorithm Of Hierarchical Closeness Centrality

While social networks are extremely popular,how to acquire and maintain some key and important attributes in social networks has become a very meaningful job.Generally speaking,most of the work abstracts the social network into a dynamic graph model for structural analysis.In graph theory and network analysis,centrality is an indicator to determine the importance of nodes in the network,and it also is a quantification of the importance of nodes.These centrality metrics were initially applied in social networks,and then they were extended to the analysis of other types of networks.In the process of analyzing social networks,one of the most basic tasks is to distinguish those objects in a community that are more influential than others,so as to help researchers analyze and understand the role of actors in the network.To complete this analysis,these people and the connections between people are modeled into a network graph.The nodes in the network graph represent people,and the edges between nodes represent the connections between people.Based on the established network structure diagram,a series of centrality measurement methods can be used to calculate which individual is more important than others.Centrality is a commonly used concept in social network analysis(SNA)to express the degree of a point in a social network or the center of a person in the entire network.This degree is represented by numbers.It is called the centrality(that is,by knowing the centrality of a node to understand the concept of judging the importance of this node in this network).Hierarchical Closeness Centrality is aimed at the small world characteristics of the social network itself.By introducing the feature of hierarchy,the enhancement highlights the step-by-step links in the social network and concentrates on the important knots.The characteristics of the point,so that the traditional close proximity to the center can better describe the structural characteristics of the analyzed social network.At the same time,in order to reduce the time cost of calculating the step type close to the centrality,a distributed algorithm corresponding to it is proposed.The test results show that the Hierarchical Closeness Centrality can effectively improve the structure of the described social network to make it more in line with its own small world characteristics,and to identify the most influential nodes and specific problems found in the community.It shows better results than the traditional close to the central.At the same time,the proposed parallel computing algorithm also shows less time overhead than the basic method.