A family of agent based models

With the increasing need to thwart the efforts of global terrorism and for the United States to effectively protect its citizens and infrastructure from attack, analysts must be provided with an improved set of tools and capabilities to evaluate, and predict, potential threats. We describe preliminary efforts associated with development of an approach based upon the concept of a family of agent-based models to represent, evaluate and predict potential terrorist activity. This system of models will be employed to search a large information repository (Infosphere) of intelligence, public record, and financial, etc. data to identify suspicious (or potentially suspicious) activity. The approach will rely heavily on concepts associated with modeling field theory (MFT) to identify data relationships. As part of the work associated with this effort, a simple example will be provided to identify the general approach and to aid in evaluation of its feasibility for the large-scale problem associated with counter-terrorism.