The Application Research of Social Networks Community Detection with Correlation Coefficients

DOI

10.2991/icence-16.2016.112How to use a DOI?

Keywords

algebraic connectivity; matrix reordering; laplascian matrix; correlation coefficients; social networks

Abstract

In complex network graph, cut model based on edge centrality doesn't apply to overlapping community detection by minimizing the algebraic connectivity of complex networks. The problem can be resolved by calculating node Correlation coefficients. It cuts one edge at one time, which is the fastest decline in the algebraic connectivity, until to divide into two communities. When components of fielder vector is less than threshold level 'a' and the difference of node adjacent edges which belong to two groups is less than 2, correlation coefficients is calculated. This node's community can be detected by the correlation coefficients. We concluded that elapsed time by ordering matrix is less than before. The advanced cut model can be used in overlapping community detection with higher efficiency and accuracy.

Copyright

© 2016, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Fuqiang Zhao
AU  - Guijun Yang
AU  - Enjun Xing
AU  - Li He
PY  - 2016/09
DA  - 2016/09
TI  - The Application Research of Social Networks Community Detection with Correlation Coefficients
BT  - Proceedings of the 2nd International Conference on Electronics, Network and Computer Engineering (ICENCE 2016)
PB  - Atlantis Press
SP  - 588
EP  - 593
SN  - 2352-538X
UR  - https://doi.org/10.2991/icence-16.2016.112
DO  - 10.2991/icence-16.2016.112
ID  - Zhao2016/09
ER  -