|Acquisition Program: ||Intelligence Analysis System, ACAT IV, PM Intel, Marine Corps Systems Comma|| Objective: ||The building intent topic will develop an analysis capability to generate inferences concerning the location of man-made structures that support the continuation of serial crime in an area. Inference development will use available data on the locations of past events, geographic information and any additional available data about facilities in the area of interest. Predictions will be based on decision models that assume that the criminal is optimizing a utility function when selecting both supporting facilities and attack locations. The capability to be developed is relevant to the Global War on Terror (finding safe houses, weapons caches, etc.) as well as to serial crime such as drug dealing.
|| Description: ||In support of repetitive crime, such as the insurgent attacks associated with the war on terror, responsible hostile actors need to leverage supporting facilities. To sustain their hostile behavior bad actors (insurgents, serial killers, drug dealers, etc.) need places from which they can train, plan, store material or simply sleep. The urban clutter affords considerable concealment for the actors that we must be capture. The ease with which their supporting facilities can blend into the clutter further complicates the task of finding civilian distributed hostile networks. Novel capabilities are needed to aggregate and classify facilities, develop facility network diagrams and ontologies that relate details about a facility (e.g. location, size, floor plans, intended use, references in human intelligence or open source data, etc.) to hostile actions via the decision preference factors of the hostile actors. A facility network diagram, for example, would connect facilities used to plan with facilities used to store to facilities used for temporary housing. Facility classification should be based on both geographic profiling and a developed understanding of what building features make the best supporting facilities to a reasoning hostile actor. The analysis engine that results from this effort will be dynamic, allowing new data sources to be continuously used to improve predictions and detect changes in past patterns. The Building Intent project will provide near real time data prediction tools that will aid in shutting down facilities that support repetitive hostile action.
The challenges in this project revolve around the following:
• Formulating a group of topographically specific data sets that will be required for geographical profiling within various urban environments.
• Formulating data layers on other facility characteristics of interest (e.g. size, floor plan, intended use, references in human intelligence or open source data etc.)
• Developing facility aggregation and classification algorithms
• Developing a capability to derive facility network diagrams
• Development of an ontology based decision model that enables the locations of supporting facilities to be inferred based on a causal understanding.
• Defining a method for assigning a measure of strength for all predictions.
• Creating algorithms that can learn form previous predictions and/or regional differences in insurgent tactics, techniques and procedures
|| ||PHASE I: Conduct research to evaluate the viability of finding facilities that support repeated hostile action in an urban environment. Identify good data sets that can be used for algorithm verification. Develop algorithms that can predict the location of supporting facilities from observations of hostile actions, detailed geographic data and detailed information about facilities in the area of interest. Develop an approach to handle regional differences and dynamic predictive model updates. Submit a report covering the approach, design and results.
|| ||PHASE II: Develop a working prototype for the phase 1 capability and demonstrate its capability against a relevant data set. Extend the capability of the prototype to include the generation of a causal ontology and network facility diagrams. Deliver and demonstrate the working prototype. Deliver a final report documenting the performance and capability.
|| ||PHASE III: Demonstrate that the products developed under phase I and II can be applied to civilian serial crime. Provide Building Intent documents and prototypes to many DoD and contractor test facilities.
PRIVATE SECTOR COMMERCIAL POTENTIAL/|| ||DUAL-USE APPLICATIONS: Topic has direct relevance to civilian law enforcement in that it will develop a more accurate serial crime prediction capability. The offeror will fuse geographic profiling with other available information to increase the accuracy in the resulting prediction.
|| References: ||1. http://www.ecricanada.com/geopro/index.html
|Keywords: ||Profiling, Insurgents, IED, Networking, Space-time predictive analysis, geographic profiling |