Rewriting general conjunctive queries using views

Abstract

The problem of rewriting queries using views has important applications in data integration, query optimization, and physical data independence maintenance. Previous researchers have proposed rewriting algorithms for queries and views that are Datalog programs or conjunctive queries with arithmetic comparisons such as x < y and y ≥ 10. We present a method for finding rewritings of general conjunctive queries, i.e, conjunctive queries with arbitrary built-in predicates, using views. Our method also has advantages over previous algorithms when there are no built-in predicates or when the built-in predicates are conjunctions of arithmetic comparisons. In particular, our method finds strictly more rewritings than the MiniCon [PL00] and the Shared-Variable-Bucket [Mit01] algorithms and tends to be more efficient when the built-in predicates of the query involve only distinguished variables. It finds all rewritings that can be found by the Bucket [LRO96] algorithm in most practical cases, and more efficiently. It finds maximum rewritings in several special cases.