The Knowledge Dictionary: A Relational Tool for the Maintenance of Expert Systems

In this paper, we discuss the development and use of a Knowledge Dictionary, a tool to facilitate the documentation and maintenance of rule based expert systems. The Knowledge Dictionary may be used to record heuristics and their component parts, facts and rule actions,utilizing the relational data model to store the heuristics in a data form rather than executable code form. Thus knowledge based systems can be documented in the same way as conventional systems. Use of the relational model allows examination of the knowledge in a knowledge base to be completely flexible, in contrast to conventional expert system tools. This flexibility is essential if the Feigenbaum bottle-neck is to be relaxed and if maintenance is to proceed in an orderly and logically consistent fashion. This paper discusses the use of the Knowledge Dictionary to redevelop GARVAN-ES1, a medical expert system for the automatic clinical interpretation of laboratory reports.