From Santa Fe Institute Events Wiki
Statistical Inference for Complex Networks Workshop, December 3-5, 2008, Santa Fe NM
Jake Hofman (homepage)
Community detection: model fitting, comparison, and utility
Much recent work has focused on community detection, or the task of identifying sets of similar nodes from network topology. Underlying this work is the implicit assumption that inferred communities inform node attributes or function in a meaningful and useful sense. We investigate these ideas by phrasing community detection as Bayesian inference, which provides a scalable and efficient algorithm for fitting and comparing network models, and applying the resulting algorithm to a university e-mail data set that includes both topology (who e-mailed whom) and node attributes (age, gender, academic affiliation, etc.). We study the relationship between the identified topological communities and the node attributes and discuss implications for community detection as a tool for network analysis.