Case authoring: From textual reports to knowledge-rich cases

Abstract

SmartCAT is a Case Authoring Tool that creates knowledge-rich cases from textual reports. Knowledge is extracted from the reports and used to learn a concept hierarchy. The reports are mapped onto domain-specific concepts and the resulting cases are used to create a hierarchically organised case-based system. Indexing knowledge is acquired automatically unlike most textual case-based reasoning systems. Components of a solution are attached to nodes and relevant parts of a solution are retrieved and reused at different levels of abstraction. We evaluate SmartCAT on the SmartHouse domain looking at the usefulness of the cases, the structure of the case-base and the retrieval strategy in problem-solving. The system generated solutions compare well with those of a domain expert. © Springer-Verlag Berlin Heidelberg 2007.