Using Graphs to Explore Communication Networks
Chris Volinsky, (AT&T Labs-Research), email@example.com
When studying transactional networks such as telephone call detail data, credit card transactions, or web clickstream data, graphs are a convenient and informative way to represent data. When the graph edges represents actual communications between transactors, the graph can then be mined to find communities of communicators, for the purpose of detecting fraud cells or marketing segments. Through a combination of visualization, graph theory algorithms, and statistical analysis, we can learn things from the graph that we could not have discovered otherwise. In this talk I will introduce some graphs from our communications networks, and discuss how we have used these tools to find interesting communities.