Verification and validation of rule-based expert systems

• Other Authors: Smith, Suzanne.
• Format: Others
• Language: English
• Subjects: Computer Science; Artificial Intelligence
• Online Access: http://purl.flvc.org/fsu/lib/digcoll/etd/3087561

Verification and validation are essential to providing quality assurance for expert systems. The same level of quality assurance is expected of expert systems as has been demanded in conventional software. Verification and validation provide a systematic, comprehensive approach to quality assurance for conventional software. Such quality assurance is lacking in the development of expert systems. Current practice in verification and validation consists of the informal, haphazard testing of a few test cases. A lesson from the software crisis of conventional software is that quality can not be measured by such testing. Therefore, a systematic, comprehensive approach to verification and validation is necessary if quality assurance is to be provided for expert systems. === A methodology for the verification and validation of rule-based expert systems is developed in this research. The research objectives are to define verification and validation for expert systems, to delineate the activities required for thorough verification and validation of rule-based expert systems, and to develop a comprehensive set of techniques and tools for validation of rule-based expert systems. To this end, the contributions of this research are the recommendations for a systematic, comprehensive approach to the verification and validation of rule-based expert systems and the development of a complete set of techniques and tools for effective validation of rule-based expert systems. === Source: Dissertation Abstracts International, Volume: 52-03, Section: B, page: 1557. === Major Professor: Abraham Kandel. === Thesis (Ph.D.)--The Florida State University, 1991.