Evaluating a case-based discovery system: a case-study in content-oriented evaluation

Abstract

Research in case-based reasoning has explored the ideas of retrieving and adapting previous experience in the form of cases. The retrieval process uses an index vocabulary which attempts to capture content and context of previous cases. Current approaches to empirical evaluation of case-based reasoning systems often do not consider the crucial role which a content-oriented index vocabulary plays in the various stages of the case-based reasoning process. We will address this issue by describing an evaluation of our case-based discovery system IULIAN. Based on this description, we will propose a general schema for content-oriented evaluation of case-based reasoning systems.