Targeting Critical Information Requirements with Robust Knowledge

We propose to target critical information requirements with an innovative extensible agent architecture that employs multiple levels of knowledge discovery and information fusion. The proposed system concept, Informant, will employ a combination of machine learning-based bootstrapping techniques and end-user accessible ontology development tools to facilitate the development and maintenance of background knowledge. This background knowledge will enable an analogy based CCIR capture process in which Informant infers information requirements from a case base of historic combat situations and a description of the unique combat situation that gave rise to these requirements. Informant will further apply its background knowledge to translate these CCIRs into specific RFIs and other machine-actionable guidance. Once codified, CCIRs will drive a knowledge discovery process that may involve anything from straightforward data retrieval to intricate information extraction and inference depending on the form of requirement and the availability of the information resources at the time. Once identified, critical information will be disseminated in a concise but unpackable form compatible with mobile devices. Phase I research and development of a limited prototype will provide a solid foundation for the complete implementation of Informant in Phase II and its eventual commercialization.