Title: Community structure in complex networks
Speaker: Mark Newman
Speaker Info: University Of Michigan
Brief Description:
Special Note: More current information may be available at Plan-it Purple
Abstract:
Many networks, including technological, social, and biological networks, divide naturally into communities. For instance, communities on the world wide web might represent groups of pages on related topics, communities in a biochemical network might represent functional groups. Traditional methods for detecting community structure in networks, like spectral partitioning and hierarchical clustering based on structural equivalence measures, give poor results when applied to such real-world problems. I will present number of new algorithms and methods for detecting community structure, particularly methods based on betweenness measures and methods based on modularity optimization. Examples of the application of these methods to a variety of network datasets will be given, including collaboration networks, acquaintance networks, and food webs.Date: Friday, March 12, 2004