Natural arising and evolution of community structures in natural and social networks has been explained as a result of topological relationships among nodes in the same network, and many studies in this field have revealed that it is possible to derive information about the community decomposition of a network just by examining its structure. The most used metric for this kind of analysis is the so-called "modularity" [12][11], which expresses the quality of a candidate community decomposition of a network. Despite its popularity, modularity is hard to be optimized [2] and algorithms for communities discovering based on modularity optimization are practically unfeasible for large networks. On the other hand, methods for community uncovering based on locally evaluated metric are very fast [7]. In this paper we propose the use of a parallel implementation of the local metric based method for community discovering proposed in [9] and the use of the overlapping modularity function [13] to evaluate the best partition. All measures reported in this paper are obtained running our implementation within a grid computing environment. © 2009 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Fiumicello, D., Longheu, A., & Mangioni, G. (2009). Discovering community structure on large networks using a grid computing environment. In Studies in Computational Intelligence (Vol. 207, pp. 63–71). Springer Verlag. https://doi.org/10.1007/978-3-642-01206-8_6
Mendeley helps you to discover research relevant for your work.