Probabilistic Evidential Reasoning for Collecting Signals Intelligence Data

The goal of this project is to develop a scalable software platform for SigInt collection management and data fusion. This platform will help provide SigInt analysts and users with effective access to the right information at the right time while helping to reduce information overload. The platform is designed to support a cognitive style of information processing and data fusion. The computational methods are cognitive in the sense that they implement a logico-mathematical process that is flexible, context sensitive, and hypothesis-driven, emulating the way in which human experts assimilate and interpret arriving data, progressively transforming data into information and understanding, while gracefully adapting or revising old hypotheses as needed based on newly detected events and conditions. The platform uses information framing and multi-layer fusion to organize collected information, to group, filter, and fuse signals intelligence data, and to organize the presentation of data to users and analysts. The framing and fusion algorithms provided by this toolset will be used to support the work and enhance the effectiveness of SIGINT analysts as they assess the relationships between detection events to discover patterns and trends, improving the accuracy and efficiency of analyst detection, identification, and geo-location of emerging threats.