Visualizing Online System For Clusting Network

In recent years,a large number of papers on the theory and application of network science have been published,and network research has attracted more and more people's attention.In the process of network analysis,displaying and operating the network through data visualization can intuitively reflect the structure of the network and help researchers better mine the network information.At present,most of the network analysis tools are single-machine software or application systems based on C/S structure.With the development of Web applications,browser-based online visualization technology has attracted people's attention for its advantages of good cross-platform,strong interaction and convenient deployment.This paper designs and implements an online visualization platform for network clustering.Considering the large scale of complex networks,in order to better analyze the network,it is necessary to compress the network by clustering algorithm.After clustering,the network structure is relatively simple and convenient for analysis.Based on this strategy,the main research contents of this paper are as follows:the Walktrap clustering algorithm based on random walk is studied and evaluated;the simplified strategy of network compression visualization based on Walktrap algorithm is designed and implemented;and the online visualization platform of network clustering based on ECharts framework is designed and implemented.The online visualization platform of network clustering consists of network input module,network clustering module and visualization module.It realizes the visualization of clustering network,as well as the functions of community drill-down inspection,network clustering quality analysis,distance query,community internal topology information query,network filtering and node attribute screening.The network clustering quality analysis function supports users to view the modularity and contour coefficient.The distance query function allows users to view the distance between nodes and communities.The internal topology information query includes the query of the number of nodes in the community,the diameter of the community and the core nodes of the community.The network filtering function mainly carries out on the nodes and edges in the network through the degree of nodes and the boundary weight value.Delete,and the node attribute filtering function allows users to filter and view nodes according to their attributes.Finally,the tool is tested to verify its effectiveness and practicability.