Kuo-Tay Chen
National Taiwan University
Ronald M. Lee
Erasmus University Rotterdam
Abstract: The review and evaluation of internal accounting control systems has always been a major task to auditors and management due to professional, legal, and economic concerns. Though there are several tools developed to help auditors with the mental modeling and reasoning process, these tools still have some limitations. This study has proposed and validated a pattern recognition approach for automatically identifying fraud potentials exposed by an internal accounting control system. A prototype knowledge-based system has been developed to validate the approach. The knowledge-based system takes a model of an internal accounting control system as input, matches it against audit patterns, and identifies fraud potentials associated with matched audit patterns. To facilitate the development of the knowledge-based system, we have dealt with the issues of representational formalism, knowledge base derivation, and pattern matching.
Back to Program
