Engineering for Artificial Intelligence Software

摘要: Abstract : Rule based systems are being applied to tasks of increasing responsibility. This report focuses on techniques for the verification and validation these systems. Conventional software quality assurance depends availability requirements specification documents. For rule systems, there generally none because capabilities evolve through a development process that is partly experimental in nature. testing considered; however, such do not carry over absence errors. Methods proving consistency examined. These methods require rules be viewed declaratively, which may too much simplification. A semantics term rewriting developed. Standard tests confluence cannot converted firing can depend absence, as well presence, elements working memory. Finally, we consider deductive