An indexing vocabulary for case-based explanation

摘要: The success of case-based reasoning depends on effective retrieval relevant prior cases. If is expensive, or if the cases retrieved are inappropriate, and adaptation costs will nullify many advantages from experience. We propose an indexing vocabulary to facilitate explanations in a casebased explanation system. we consider anomalies (conflicts between new situations expectations beliefs). Our groups according type information used generate beliefs that failed, how failed. argue by using this characterize anomalies, retrieving were built account for similarly-characterized past system can restrict likely be relevant. In addition, organize general strategies suggest paths novel situations.