Open AccessCommunity Detection Research Based on Line Graph
Author(s) -
Chun Gui
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1881/3/032051
Subject(s) - community structure , spectral clustering , computer science , graph , cluster analysis , spectral graph theory , complex network , clustering coefficient , line graph , similarity (geometry) , graph theory , algorithm , theoretical computer science , laplace transform , pattern recognition (psychology) , data mining , artificial intelligence , mathematics , combinatorics , image (mathematics) , voltage graph , world wide web , mathematical analysis
Community structure is an important property of complex networks, in which network nodes are tightly split into different groups. In this paper, a new overlapping community detection algorithm based on line graph is proposed. Spectral analysis based on Normal Laplace Matrix of line graph is introduced. We not only make the spectral clustering to be an overlapping algorithm but also define a new similarity between edges by spectral analysis. We test our method on three-community network and Zachary karate club network. The experimental result shows that our method can detect community structure and overlapping nodes effectively. It is worth to note that our method is against undirected and unweighted graphs.