Redundancy: an approach to the efficient implementation of semantic integrity assertions

Semantic integrity assertions accurately model the real world by defining consistent database states; however, the assertions are expensive to enforce because checking them may require numerous database accesses. The authors therefore propose a method which alleviates this expense by storing redundant data. Integrity constraints are assumed to be expressed in a relational calculus-like language with an arbitrary number of quantifiers and a general logical structure. The method chooses clustered redundant data such that the benefit of using the redundant data for testing the assertions outweighs the cost of their maintenance during updates