Abstract
Semantic integrity constraints are used to help ensure that a database accurately reflects the real world in structure and content. Existing database design methodologies, however, do not often include a comprehensive procedure for identifying such constraints and incorporating them into a design. A number of knowledge-based systems for database design have been developed with varying degrees of emphasis on, and approaches to, this task. A representative sample of these systems is reviewed for their ability to identify, represent and validate integrity constraints. In order to examine how constraints are represented, a six-dimensional model of semantic integrity constraints based on that proposed by Fernandez et al. [32] is presented. Mechanisms used by the prototype systems to implement each of the dimensions are examined. This analysis is used to identify apparently successful approaches as well as areas where further research on incorporating and representing semantic integrity constraints in knowledge-based database design systems is needed.