Recursive query processing: the power of logic

Abstract

The purpose of this paper is to show that a computational model developed in the framework of resolution provides a very adequate tool to study and develop query answering procedures for deductive databases, as well as for logic programs. As a result, we introduce an effective query answering procedure for deductive databases. To achieve our goal, we first develop techniques (applicable to general logic programs) for the construction of abstract search spaces associated with a query, and we discuss their properties. We then show how these techniques can be practically applied to the problem of answering recursive queries in a deductive database (consisting of function-free clauses). This approach has given rise to a new general-purpose procedure, termed QoSaQ, which improves on earlier general-purpose methods, in particular by its ability of incorporating a so-called global optimization technique. We show that the framework provided by QoSaQ is powerful enough to account for the best-known recursive query evaluation methods. We also produce an upper-bound to the number of tuples manipulated by QoSaQ, which improves on known upper-bounds.