Information and Decision Recommender

A general framework for course of action (COA) recommender systems consists of four primary functional areas : (1) mapping human entered COA into formalized, machine readable knowledge representation; (2) deconstructing COA decision models into relevant features or variables of interest; (3) gathering and routing intelligence related to those features to specific warfighters evaluating COA; and (4) predictive capability that leverages the routed information sets to suggest COA recommendations. With this in mind, Securboration proposes to develop a predictive capability to rank COA , referred to as COA Recommendation Services (COARS) , that fits within the Navys vision for a COA recommender. COARS is not a monolithic solution, but rather a service-based implementation that specifically addresses key technology in the COA recommender vision: predictive capability to rank COA. The novelty of our Teams approach is that we leverage seminal research in computational modeling and extend it by borrowing deep learning concepts to fuse knowledge and derive complex emergent behavior that is missed with existing reasoning techniques and algorithms. This accounts for the intricacies and dynamisms of the operational environment and overcomes the limitations of current COA recommender approaches that are based on simplistic assumptions regarding adversary objectives, limitations, and intent.