Expert Intelligent Match of Requirements and Solutions

Ex-PERT, proposed under this SBIR, will aid a human decision maker in filtering contractor proposed solutions that match requirements, reducing the time to select an appropriate solution and hence, reducing the overall system development cycle-time. The work effort will 1) develop a capability to match textual descriptions of requirements to relevant textual descriptions of contractor proposed solutions to achieve improved efficiency in the system development life-cycle and 2) implement that capability in the Ex-PERT tool. The innovations include 1) a functional grammar and ontological lexicon based on the sublanguage of the Requirements, 2) an ontology of the notion of an appropriate solution, 3) an ontology of the application domain, and 4) a functional grammar and ontological lexicon based on the sublanguage of the contractor-proposed Solutions which differs from that of the Requirements. The application domain for demonstrating Ex-PERT will be the intrusion detection area. Phase I will develop the prototype based on proven computational linguistics technology in several complex areas including 1) text extraction for both the Requirements and proposed Solutions; 2) knowledge representation techniques and ontological representations for representing the Requirements text in a normalized form as well as the notion of a good solution; and 3) inference methods to determine the suitability of the matches between Requirements and Solutions. Ex-PERT will produce a score for matching of the Requirements and Solutions based on metrics of the solutions to provide a scale for comparing possible solutions.