The development of an ontology-based expert system for corporate financial rating

Abstract

This study presents a new approach in developing an expert system to assess financial quality of an enterprise. In dealing with the complexity of financial statement analysis, this approach separates knowledge content into domain knowledge of financial statements and operational knowledge of analysts’ analytical process. The former represents the well defined relationships of accounting items of financial statements. The latter represents the reasoning processes that use the domain knowledge to appraise the quality of given financial statements. We apply ontology to model the content of domain knowledge, and use decision rules to represent operational knowledge. This system integrates Protégé, as domain knowledge base, and JESS, as operational knowledge base, into one complete expert system. The performance validity of the system was verified with financial statements from Taiwanese stock market.