Title: Community structure discovery in Facebook

Authors: Emilio Ferrara

Addresses: Department of Mathematics, University of Messina, V.le F. Stagno D'Alcontres n. 31, 98166, Italy

Abstract: In this work, we present a large-scale community structure detection and analysis of Facebook, which gathers more than 500 million users at 2011. Characteristics of this social network have been widely investigated during the last years. Related works focus on analysing its community structure on a small scale, usually from a qualitative perspective. In this study, we consider a significant sample of the network. Data, acquired mining the web platform, have been collected adopting two different sampling techniques. We investigated the structural properties of these samples in order to discover their community structure. Two well-known clustering algorithms, optimised for complex networks, have been here described and adopted. Results of our analysis show the emergence of a well-defined community structure inside Facebook, that is characterised by a power law distribution in the size of the communities. Moreover, the identified communities share a high degree of similarity, regardless of the adopted detection algorithm.

Keywords: data mining; complex networks; community mining; community detection; network analysis; social network mining; community structure; network communities; clustering algorithms; Facebook.

DOI: 10.1504/IJSNM.2012.045106

International Journal of Social Network Mining, 2012 Vol.1 No.1, pp.67 - 90

Received: 04 Apr 2011
Accepted: 30 Aug 2011

Published online: 21 Aug 2014 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article