Dynamic community identification

Abstract

Humans and other social animals interact in diverse ways. A central problem in the study of societies is identifying core communities: sets of entities among whom interactions are frequent and consistent. Membership in social groups often changes, thus making it difficult to characterize a society's community structure. In this chapter we formalize the computational problem of dynamic community identification and review computational methods that address the changing nature of community membership. We discuss in detail the dynamic community identification method by the authors which in many ways subsumes other approaches.