The use of cases of diagnostic explanations

Explanation of conclusions is considered to be one of the major advantageous features of knowledge based systems: it encompasses both the understanding of the user´s queries and explanation of the reasoning. Unfortunately, this explanatory information is not provided in most existing applications, and few systems have adequate explanation facilities. Whilst explaining, experts often refer to past experiences: in this paper, the production of explanations which explain and justify a solution with reference to past cases is presented. Case-based explanations are very valuable in diagnostic explanation. They justify conclusions of the system in referring to past experiences. The level of acceptance of these conclusions by an operator who does not always understand the reasoning steps of the system, is considerably enhanced. This is typically the case with the DIVA expert system, which is based on vibration analysis, where its conclusions must be justified by concocting an explanation which is independent of the reasoning.<>