Interactive knowledge acquisition and scenario authoring

Abstract

Encoding knowledge and authoring content in virtual reality systems typically requires trained specialists. This project, concerned with training people to identify risky situations, aims to reduce the knowledge acquisition bottleneck by allowing trainers to enter their knowledge and generate alternative scenarios whilst interacting with existing training sessions. The goal is not only to speed up the capture of knowledge and content but also to tap into (tacit) knowledge through this knowledge-in-action approach and to allow the trainee to experience and query the underlying domain knowledge. Ripple Down Rules is proposed as the knowledge representation and method as it supports incremental acquisition of rules performed by the domain expert as new situations arise. In this project we seek to capture concurrently new knowledge and cases. © 2008 Springer Berlin Heidelberg.