Seminars & Colloquia
Department of Computer Science, KAIST
"Consistency in Community Identification (Social Networking)"
Tuesday April 27, 2010 11:00 AM
Location: 1212, EB II NCSU Centennial Campus
(Visitor parking instructions)
Online social networks pose significant challenges to computer scientists, physicists, and sociologists alike, for their massive size, fast evolution, and uncharted potential for social computing. One particular problem that has interested us is community identification. Many algorithms based on various metrics have been proposed for identifying communities in networks [18, 24], but a few algorithms scale to very large networks. Three recent community identification algorithms, namely CNM , Wakita , and Louvain , stand out for their scalability to a few millions of nodes. All of them use modularity as the metric of optimization. However, all three algorithms produce inconsistent communities every time the input ordering of nodes to the algorithms changes. We propose two quantitative metrics to represent the level of consistency across multiple runs of an algorithm: pairwise membership probability and consistency. Based on these two metrics, we propose a solution that improves the consistency without compromising the modularity. We demonstrate that our solution to use pairwise membership probabilities as link weights generates consistent communities within six or fewer cycles for most networks. However, our iterative, pairwise membership reinforcing approach does not deliver convergence for Flickr, Orkut, and Cyworld networks as well for the rest of the networks. Interestingly, Orkut and Cyworld are only two out of twelve networks whose community size distribution follow power-law.
Our approach is empirically driven and is yet to be shown to produce consistent output analytically. Hood et al. report that modularity distribution has a flat landscape and partitions with close modularity values may be very different from one another. We conclude this talk with a discussion on our ongoing work and future directions.
Sue Moon received her B.S. and M.S. from Seoul National University, Seoul, Korea, in 1988 and 1990, respectively, all in computer engineering. She received a Ph.D. degree in computer science from the University of Massachusetts at Amherst in 2000. From 1999 to 2003, she worked in the IPMON project at Sprint ATL in Burlingame, California. In August of 2003, she joined KAIST and now teaches in Daejeon, Korea. She has served as TPC co-chair for ACM Multimedia and ACM SIGCOMM MobiArch Workshop, general chair for PAM, and TPC for many conferences, including SIGCOMM 2010, NSDI 2008 and 2010, WWW 2007-2008, INFOCOM 2004-2006, and IMC 2009. She is currently serving as guest editor for IEEE Network Special Issue on Online Social Networks and Journal of Network and Systems Management Special Issues on New Advances on Measurement Based Network Management. She won the best paper award in ACM SIGCOMM Internet Measurement Conference 2007 and has been awarded the Amore Pacific Woman Scientist Award in 2009. Her research interests are: network performance measurement and analysis, online social networks, and networked systems.
Host: Injong Rhee, Computer Science, NCSU