Agent/Source Characterization for exploitation of human-generated intelligence

Agent/Source Characterization (A/SC) technology is being developed to improve the effective exploitation of human-generated data. This technology is being incorporated into a prototype system for assessing and predicting activity in threat-related social networks. Classical multi-source, multi-target calibration/registration / state estimation methods are applied to the joint characterization of human information sources and their reported target/ situation state estimates/ A Structural Equation Model framework is used to integrate a source agent's measurement, inference, planning and control processes with ldquoourrdquo independent observations of agent, targets and situations. Human behavior models are embedded in this framework to characterize both human sources and human targets.