Re: [igraph] Degree of confidence that a node is member of a community

[Gábor Csárdi]

On Thu, May 10, 2012 at 12:55 PM, Tamás Nepusz <address@hidden> wrote:
>> I was wondering if there is a way to identify significance in the
>> identification of communities. Like a degree of confidence that one
>> node of the network is member of a specific community. Is there any
>> function, or is there any attribute of the class community, that gives
>> such information?
>
> No, there isn't -- as far as I know, there is no established algorithm in the 
> literature that would calculate such a confidence value, although a few 
> papers have been published recently that try to tackle this issue. You might 
> be interested in checking out the OSLOM algorithm by Lancichinetti et al -- 
> this is an algorithm that returns statistically significant communities:
>
> http://www.oslom.org/
> http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0018961
>
> Another related paper was published here:
>
> Lancichinetti A, Radicchi F, Ramasco JJ (2010) Statistical significance of 
> communities in networks. Phys Rev E 81: 046110.
>
> Best,
> T.
[...]

Also, in some sense, the actual values in the eigenvector denote the
strength of community membership in the leading eigenvector method by
Mark Newman. See
http://arxiv.org/pdf/physics/0605087v3.pdf
http://www.pnas.org/content/103/23/8577.full

Gabor