Integrating Dynamic Social Networks and Spatio-Temporal Models for Risk Assessment, Wargaming and Planning
Abstract
Most social network modeling tools are descriptive.That is, they focus on representing and visualizing entities and relationships given historic data. While such representations are valuable for understanding the structure of organizations and groups, they leave the process of projecting the entities future actions within the context of the network entirely to the user. Building on technologies from DARPAs RAID and COMPOEX programs, we are working to fill this gap by developing a system to aid analysts in projecting future activities in space, time, and socio-political dimensions. Our approach uses a generative model of a social network, related physical processes, and the environment (a geo-spatial model). The nodes (entities) in our networks have goals, resources and locations. They act, using their resources, to achieve their goals. These models generate estimates of future behavior using faster than real time simulation within the network, process, and geo-spatial models. Using such a system, an analyst can more rapidly explore questions such as What is the chain of relationships, processes and locations that sits between individual bomb parts and an IED attack? and If we crack down in a given region, where are new attacks likely to surface? Our system does not replace analysts but helps analysts more rapidly and thoroughly explore potential futures, assess risk, and plan for contingencies.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2009
- Accession Number
- AD1020794
Entities
People
- Jacob Crossman
- Robert Bechtel
- Sven Brueckner
- Van Parunak