Font Size: a A A

Research On Community Detection Algorithms In Complex Networks

Posted on:2019-12-14Degree:MasterType:Thesis
Country:ChinaCandidate:Y Q ChenFull Text:PDF
GTID:2370330566496040Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
In recent years,with the rapid development and reform of information and communication technologies,researches on complex networks have also attracted more and more attention from researchers.Community structure is an important feature of complex networks,which is of great value and significance to the in-depth study of complex networks' structure and functional characteristics.Through the community detection of complex networks,it is possible to detect the implicit group structure in the network,to reveal some structure or hidden related information contained in the real data set of constructing complex networks,which has important practical value.For the community detection in complex networks,the work of this paper is as follows:(1)A new local similarity index is proposed.This index is based on the RA index,introducting the closeness of the measured node pairs and their mutual neighbors.The improved RA similarity index can more accurately measure the similarity between nodes in the network and improve the accuracy of the similarity clustering between nodes,making the community detection algorithm based on similarity more accurately identify high quality network community structure.(2)Based on the improved RA similarity index,and combining the improved K-means algorithm,a community detection algorithm based on local similarity named CDALS is proposed.The simulation experiments are carried out on the reconstructed networks of real network data sets.Comparing with the results of representative community detection algorithm,the experimental results show that CDALS algorithm can be used to accurately divided the community structure of high quality,and also verify the validity and accuracy of the improved RA similarity index.(3)Based on the definition of local influence among node to which the node directly connected with,this paper improves the traditional label propagation algorithm from the aspects of node label update order and label update strategy,and proposes a label propagation community detection algorithm based on local influence named LPALI.Through the simulation experiment on the real network,compared with the classical community discovery algorithm,it is proved that the LPALI algorithm has the high accuracy,stability and execution efficiency,can detect the community structure of high quality and stable division quickly,and can be applied to a larger network.
Keywords/Search Tags:Complex network, community detection, similarity, K-means, influence, label propagation
PDF Full Text Request
Related items