Knowledge Representation for Military Mobility Decision-Making by Humans and Intelligent Software

Abstract

The U.S. military is constructing a Global Information Grid that provides key software services to complex networks of computers and software clients in an operational theater. Commanders’ need for accurate and timely information in support of complex decisions requires that application programs, intelligent agents and humans be able to exchange, analyze, interpret and report information to one another. While interoperability of human soldiers has traditionally been accomplished by the creation of tacit and explicit knowledge through training, construction of software applications and intelligent agents for the GIG requires a standardized vocabulary and semantically rich formalization of common sense knowledge for the various domains of operation spanned by military planning and operations. This formalization is appropriately captured in ontologies which both provide representation vocabularies and facilitate information exchange. Our recent project to define a data model and ontology for the Mobility Common Operating Picture and our ongoing work to support dynamically computed Common Maneuver Networks illustrale the knowledge engineering challenges inherent in a domain where humans have traditionally relied on tacit knowledge to evaluate information as it influences key decisions. Distinguishing concepts that are inherently relational in nature from those that represent object attributes is a key success factor.Key wordsKnowledge representationontologyartificial intelligenceformal concept analysisroute finding