Node Influence Propagation Overlapping Communities Detection And Applications Research

Complicated in the real world,any system that contains a number of individual and link can be abstracted as the network,how to explore a large overlapping community structure of complex networks is a hot topic today,a large number of overlapping community discovery algorithm was proposed and applied to all areas of large-scale complex networks,help reveal the value of the network data set.Today,the field of bioinformatics is booming,and researchers are applying community detection algorithms to protein-protein interaction(PPI)networks to help predict the function of unknown proteins at the molecular level,thus further revealing the regularity of cellular activity.In order to solve the problems of community discovery effect and universality and unreasonable community scale of the existing overlapping community discovery algorithm,this paper proposes an algorithm for discovering overlapping communities based on node influence propagation coefficient,abbreviated as NIP algorithm.Based on HIE algorithm and LPANNI algorithm,this algorithm combines node influence measurement method based on three-level neighbor and triangular structure influence measurement model to propose node influence propagation coefficient.In this way,the accuracy of central node selection is improved: the measurement method of edge distance is improved to make the result of community division more accurate;the greedy optimization strategy of community edge node is proposed to make the overlapping structure more reasonable.The results show that the NIP algorithm proposed in this paper can significantly improve the EQ and NMI values discovered by the network community compared with HIE algorithm,CES algorithm,LPANNI algorithm and LC algorithm.It is verified that the algorithm can not only present reasonable community structure,but also find overlapping structure in the network.At the same time,it is found in the research process that there is currently a lack of visualization system for knowledge map of biological network.Therefore,this paper designs and implements a visualization system for bioinformatics analysis results based on Neo4 j and Cytoscape.js.It can make knowledge map of the results discovered by biological network community for people to refer to.To present visible and queryable experimental data for later researchers and promote the application development of community discovery algorithms in the field of bioinformatics.