Facilitated Argumentation through Automatic Acquisition and Synthesis of Time-critical information (FAAAST)

As ever more data becomes available via faster, more reliable networks and advanced sensor platforms, there is an ever more persistent need for technology that will automatically find, synthesize, and contextualize data so that it may be used to support decision making. This need is especially pressing in time-critical targeting (TCT) operations. In this proposal, we present a plan to develop a platform for Facilitated Argumentation via Autonomous Acquisition and Synthesis of Time-critical information (FAAAST). The proposed FAAAST platform leverages our in-house structured argumentation engine, an ontology for time-critical information, and our in-house multi-agent query system. Decision templates in the form of argumentation networks will be developed for a range of decisions in time-critical contexts. Upon instantiation, a network will immediately cause multiple query agents to be spawned, leveraging the installed ontology, to gather information from various of information sources. As information is retrieved, the argumentation network updates the best decision option and confidence. The decision maker can interact with the network to modify its structure and weights, and drill into information sources to request more data. FAAAST’s capacity for continuous, incremental improvements in decision quality make is highly suitable as a decision support tool in the TCT domain.