Using empirical analysis to refine expert system knowledge bases

摘要

seek is a system which gives interactive advice about rule refinement during the design of an expert system. The advice takes the form of suggestions for possible experiments in generalizing or specializing rules in a model of reasoning rules cited by the expert. Case experience, in the form of stored cases with known conclusions, is used to interactively guide the expert in refining the rules of a model. This approach is most effective when the model of the expert's knowledge is relatively accurate and small changes in the model may improve performance. The system is interactive; we rely on the expert to focus the system on those experiments that appear to be most consistent with his domain knowledge. The design framework of seek consists of a tabular format for expressing expert-modeled rules and a general consultation system for applying a model to specific cases. Examples are given from an expert consultation system being developed for diagnosing rheumatic diseases.