Скачать или смотреть Machine Semantics in Training Scenario Development (Event Generation and Scenario Bending)

Download the presentation: https://ext.sharepoint.ctc.com/ctcCom...

Training scenario development — whether for simulation based training or live exercise training, whether for individual trainees, small groups, or large groups — is most effective when the events and entities in the training scenario accurately reflect the details of the operation for which training is intended to engender readiness. Simply put, training for an operation should be a lot like the operation being trained for. Good training scenario development is typically quite labor intensive and involves the "labor" of very knowledgeable personnel. We have developed a decision support capability to significantly reduce human labor in event generation and scenario bending for training scenario development. Training scenario development can be accomplished from scratch, starting with a blank sheet of paper and conceptualizing the entire scenario to accomplish all training objectives. Conceptualizing the entire scenario includes developing OPFOR entities (individual to group), intent, methods, capabilities, TTPs, and finally activities (Event Generation). We describe a system that enables improved combat/military training by computationally building various events in an environment of combat and/or military training that is more realistic because it utilizes ontology based understanding of the real data from existing and currently utilized sources which are unclassified and typically located in commonly used battle command systems and databases. The capability can help save organizational training funds and man hours by negating the need to alter millions of classified training data/reports for training programs. Building upon RIPPLE's message ontology, which generates realistic messages of multiple types commensurate with the type of training required, is essential to current military training environments.

Dr. J. Wesley Regian is the DARPA Program Manager at Concurrent Technologies Corporation (CTC). As Principal Knowledge Engineer, he is responsible for cognitive science, knowledge modeling, and knowledge-based software engineering. He has published and presented over 100 peer-reviewed academic papers on learning, performance, and instructional technology. Among these, he published the first formal empirical evaluation of Virtual Reality as an instructional medium.