Parallel Processing Letters Vol. 22, No. 1 (2012) 1240003 (11 pages) c  World Scientific Publishing Company DOI: 10.1142/S0129626412400038 LOCATING COMMUNITIES ON REAL DATASET GRAPHS USING SYNTHETIC COORDINATES * HARRIS PAPADAKIS Department of Applied Informatics and Multimedia Technological Educational Institute of Crete, 71500 Heraklion, Crete, Greece adanar@epp.teicrete.gr COSTAS PANAGIOTAKIS Department of Commerce and Marketing Technological Educational Institute of Crete, 72200 Ierapetra, Crete, Greece cpanag@staff.teicrete.gr PARASKEVI FRAGOPOULOU Department of Applied Informatics and Multimedia Technological Educational Institute of Crete, 71500 Heraklion, Crete, Greece fragopou@ics.forth.gr Received September 2011 Revised November 2011 Communicated by Guest Editors ABSTRACT One of the fundamental problems in social networking with a lot of potential applica- tions is to detect effectively the communities that are created by the users’ interaction. Other applications, such as finding web communities, uncovering the structure of social networks, or even analyzing a graph’s structure to uncover Internet attacks are equally as important. All these problems converge to a common goal, the development of flex- ible and efficient local community detection algorithms. In this paper, we demonstrate the performance of an algorithm that uncovers the entire community structure of a net- work, based solely on local interactions between neighboring nodes and a distributed clustering algorithm. The proposed algorithm, named VCD, is based on the distributed computation of a synthetic coordinates for each graph node. Experimental results and comparisons with another method from the literature (Lancichinetti et al.) are presented. The algorithm is also tested on two real dataset graphs from the SNAP: Stanford Large Network Dataset Collection. In all cases the experimental results demonstrate the high performance of our algorithm in terms of accuracy to detect communities, and its com- putational efficiency. Keywords : Community detection, network coordinates, clustering. * This project is implemented through the Operational Program “ARCHIMEDE III: Education and Lifelong Learning” (project P2PCOORD) and is co-financed by the European Union (European Social Fund) and Greek national funds (National Strategic Reference Framework 2007 - 2013). 1240003-1