Mathematical Methods in Counterterrorism

Abstracts

Bayesian Modeling of Group Structure in Dynamic Networks

Jim Ferry
Senior Analyst
Metron, Inc.

Abstract:

This talk generalizes the framework used for finding groups on static networks to produce a Markov process model for the evolution of group membership in a dynamic network exhibiting multiple link types. An efficient simulation of the Markov process model has been developed, as has an exact inference method for exploiting temporal network data in an optimal manner in order to track the group membership of all vertices in a network. An example of inference in a simple scenario is presented. The results demonstrate how both positive information (changes in network structure) and negative information (periods of no change) may be combined to track group membership optimally. The goal of this research is to develop a general capability to determine the structure and evolution of relationships between people or other entities based on data about the history of their interactions.

Back to Abstracts