INTEGRITY AND RULE CHECKING IN NONMONOTONIC KNOWLEDGE BASES

Anomalies such as redundant, contradictory or deficient knowledge in a knowledge base indicate possible errors. Various methods for detectin g such anomalies have been introduced, analyzed and applied in the pas t years, but they usually deal with rule-based systems. So Far, little attention has been payed to the verification and validation of nonmon otonic knowledge bases, although there are good reasons to expect that such knowledge bases will be increasingly used in practical applicati ons. This paper discusses how classical verification methods may be ap plied to detect some anomalies in nonmonotonic knowledge bases. These anomalies are first described in a formal way, and then generic verifi cation methods to detect them are presented.