|
From: | Gábor Csárdi |
Subject: | Re: [igraph] Community detection based on conductance |
Date: | Tue, 6 May 2014 17:11:56 -0400 |
Hi,I am performing community detection on citation network graphs (~20k nodes). It seems like all (most?) community detection algorithms are based on modularity which according to this paper (http://dl.acm.org/citation.cfm?id=2350193) is a bad idea. They propose conductance (or e.g. triangle participation ratio) as a metric to optimize for communities. In particular I am interested in a score for maximum community saliency (or e.g. minimum conductance cut).Does iGraph have such capabilities? I could find anything about conductance in the docs.I believe the Stanford SNAP library has similar functionality (C++) but I would prefer staying with Python if possible.Any comments and ideas are very welcome!
Thanks,Tim
_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help
[Prev in Thread] | Current Thread | [Next in Thread] |