Resumen de Explanations in simulation based learning
Tariq M. Khan, Keith E. Brown, R. Roy Leitch
A methodology for defining multiple domain models is presented that describes structural knowledge about a process plant and human expertise in operating the plant. An analysis of how industrial training and learning is effected by this approach is presented. Central to the instructional strategy if a facility to generate explanations from the heterogeneous knowledge in the multiple domain models, and a theoretical model of explanation is described that contains a variety of explanations produced from various combinations of the multiple domain models. Each explanation is generated from several target pieces of knowledge called knowledge fillers, e.g., expert action and error type. The theoretical model has been implemented as a prototype explanation subsystem, which is being integrated into the authoring shell for intelligent Explanation and Training with Simulation (EXTRAS). An implementation of the EXTRAS toolkit is described with an example of possible kinds of explanation produced.
