News Archives

RSS Feed

Divisionality: A measure of fine-granularity community structure

September 15, 2006

  • Date: Friday, September 15, 2006 
  • Time: 1 pm — 2:15 pm 
  • Place: ME 218

Todd Kaplan Department of Computer Science, UNM

Abstract Within the area of complex networks, there has been recent interest in the detection of community structure. In this context, a community refers to a group of densely connected vertices. Community structure algorithms aim to identify communities of nodes within a network. As a means for quantifying the success of a given community partitioning, a statistical measurement known as modularity was introduced. In this talk, I introduce an alternative measurement called divisionality. In comparison to modularity, when optimized on a given graph, this new measurement identifies finer-grained structure. I will compare the two measurements on networks of known structure and then explain how community structure plays a role in the larger scope of my research involving financial markets.

Bio Todd Kaplan is a Ph.D. student in Computer Science at the University of New Mexico. He received his undergraduate degree in Biology at Brown University and holds graduate degrees in Computer Science (Cambridge University, UK) and Applied Mathematics (Oxford University, UK). He is a member of the Adaptive Computation group at UNM, led by Professor Stephanie Forrest.