GitHub - sedjokas/PercoMCV-Code-source: PercoMCV is a novel approach for the community detection in social networks. This approach (called PercoMCV) is based on a two-step algorithm. The first step consists on the use of the clique percolation method and the second step, based on the output of the first one, measures the influence of network nodes (in order to reduce the number of unclassified nodes) by using the Eigenvector Centrality method. The approach is then tested on artificial and real networks and compared to other community detection algorithms.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
Article__ANT_14_6807.pdf		Article__ANT_14_6807.pdf
PercoMCV.py		PercoMCV.py
readme.txt		readme.txt

Repository files navigation

Please read this file if you want to run this code source

To run this code, Networkx package is required
In oder to bypass some isues that may be related to the hadware, we recommand to install Anaconda
which comes up with some interested package already intalled such as networkx
Purposely for illustration we consider the graph Graphe. However, Graphe can be of any format supported by Networkx

Note: The code for the implementation of the first step is a short form of the k_clique_communities from Conrad Lee and Aric Hagberg

From authors

About

PercoMCV is a novel approach for the community detection in social networks. This approach (called PercoMCV) is based on a two-step algorithm. The first step consists on the use of the clique percolation method and the second step, based on the output of the first one, measures the influence of network nodes (in order to reduce the number of unc…

Readme