Spatial and Temporal Threat Prediction System (STTPS)

We propose to develop a Spatial and Temporal Threat Prediction System (STTPS) to be used to predict future terrorist incidents. The system will not only predict the estimated time for the attack, but will also predict the location for the attack and will opportunistically include other threat relevant data such as type of attack, responsibility for the attack, and so on. This work will build solidly on prior work at KBSI in the areas of threat detection, data mining, information fusion, knowledge discovery, and knowledge based agents. A major challenge for this type of effort is the necessity to reconcile the information from a variety of data sources. Information may come from news feeds, intelligence data, sensor readouts, private communications and many other sources. The solution that we are proposing takes advantage of previous tools developed at KBSI for the data mining/data cleaning/information extraction/threat detection domains, enabling us to concentrate on the novel parts of the problem: terrorist event prediction. BENEFITS: The benefits that accrue from this work are quite clear: we will be able to stop terrorist incidents before they happen. Intelligence analysts will be able to usurp projected terrorist events, instead of waiting to respond. The mere presence of such a system will mitigate many types of terrorist threats.