Artificial Intelligence
Constraint logic programming languages
Communications of the ACM
Why Use a Unified Knowledge Representation?
Proceedings of the 14th International conference on Industrial and engineering applications of artificial intelligence and expert systems: engineering of intelligent systems
Knowledge Base Maintenance through Knowledge Representation
DEXA '01 Proceedings of the 12th International Conference on Database and Expert Systems Applications
A rigorous approach to knowledge base maintenance
IEA/AIE'2003 Proceedings of the 16th international conference on Developments in applied artificial intelligence
| Hi-index | 0.00 |
The aim of this paper is to show a method that is able to detect a particular class of semantic inconsistencies in a rule-based system (RBS). A semantic inconsistency is defined by an integrity constraint. A RBS verified by this method contains a set of production rules, and each production rule comprises a list of arithmetic constraints in its antecedent and a list of actions in its consequent. An arithmetic constraint is a linear inequality defined in the real domain that includes attributes, and an action is an assignment that changes an attribute value. As rules are allowed to include actions of this kind, the behaviour of the verified RBS is non-monotonic. The method is able to give a specification of all the initial fact bases (FB), and the rules from these initial FB that would have to be executed (in the right order) to cause an integrity constraint to be violated. So, the method builds an ATMS-like theory. Moreover, the treatment of arithmetic constraints is inspired by constraint logic programming.