A Collaborative Peer-to-Peer Approach to Multi-Source Threat Assessment and Prediction

The proposed innovation, entitled CiTAPS (Collaborative Intelligent Threat Assessment & Prediction System) combines collaborative software agents, intelligent machine and peer-to-peer technologies to support extensible, interoperable, and fully automated real time indications and warning extraction and dissemination. It provides users with on-demand multi-source data retrieval and fusion, overt threat notification for terrorist activities, and covert threat detection and notification through remote data source monitoring and analysis. Using multiple intelligent software agents deployed to a variety of data sources within a specific domain, CiTAPS is capable of retrieving, fusing, analyzing, and extracting anomalies from multiple sources of heterogeneous data for integration into a Common Relevant Operational Picture (CROP). These agents are able to collaborate within their own domain using standards-based communications for agent frameworks. Our team's approach is to build upon previous research findings on peer-to-peer technologies, leverage those against the Homeland Defense's Intelligent Agent problem, and create a formal standards-based architectural design and an interactive visualization tool outlining the major concepts of the proposed system.