Lee Gye Sung

The Transactions of the Korea Information Processing Society (1994 ~ 2000), Vol. 3, No. 7, pp. 1781-1791, Dec. 1996
10.3745/KIPSTE.1996.3.7.1781,   PDF Download:

Abstract

A number of research issues such as knowledge acquisition, inferencing techniques, and knowledge refinement methodologies have been involved in the development of expert systems. Since each issue is considered very complicated, there has been little effort to take all the issues into account collectively at once. However, knowledge acquisition and inferencing are closely related because the knowledge is extracted by human experts from the inferencing process for solving a specific task or problem. Knowledge refinement is also accomplished by handling problems caused during the inferencing process of the system due to incompleteness and inconsistency of the knowledge base. From this perspective, we present a method by which software platform is established in which those issues are integrated in the development of expert systems, especially in the domain where the domain models and concepts are hard to be constructed because of inherent fuzziness of the domain. We apply a machine learning technique, conceptual clustering, to build a knowledge base and rule models by which an efficient inferencing, incremental knowledge acquisition and knowledge refinement are possible.

Statistics
Cite this article