Research On The Algorithms Of Community Detection And Visualization In Complex Networks

Community structure is an important feature in many real-world networks.Many methods and algorithms for identifying communities have been proposed and have attracted great attention in recent years.Research community discovery algorithm,help to understand the network topology characteristics,found the relationships between the network nodes,and has important practical value in the fields:personalized information recommendation of the network,the electronic commerce,the safety assessment of the Internet culture,and so on.This article mainly aims at the community characteristics and structure of the complex network research,comparative study on various existing community discovery algorithm,on the basis of these,and in view of the existing problems,put forward a novel community detection algorithm,and designed a visualization solutions of network community detection.Main contributions of the dissertation are as follows:This paper analyzes,compares and summarizes the existing community detection algorithms.Based on the theory of complex network,this paper analyzes the existing typical community detection algorithms,including the community detection algorithms for overlapping and non-overlapping communities.The advantages and disadvantages of these algorithms are analyzed from multiple perspectives.It is found that there are two problems in the existing algorithms:firstly,it depends on some prior information to some extent;second,the computational complexity is high,difficult to apply to large-scale network.At the same time,the result of the community detection is too and the availability is limited.In view of the defects of existing algorithms,this paper proposes a community detection algorithm which based on the connection strength CDCS.The innovation of this algorithm is to use the connection strength between nodes to divide the nodes into different communities.Specifically,it uses the weak node hypothesis to determine the community that the node belongs to.Compared with the existing scheme,this algorithm has the advantages of simplicity,accuracy and low computational complexity.The accuracy and efficiency of CDCS algorithm in different networks are verified by using real network data and LFR standard data set.For the visualization problems of network,a visualization model of D-Treemaps network community was established in this paper.This model is a network visualization model based on multi-dimensional data representation,which can better present the effect of community discovery to users.The model uses a dynamic Treemaps visualization method to enhance the presentation of online network information,which can not only express the multi-dimensional information,but also show it in a dynamic way.Through fast and effective information navigation,the model enables users to view complete information modules within a single page and quickly lock specific information items.The real network data is used to analyze the model and prove the feasibility of the model.