On the Rewriting and Efficient Computation of Bound Disjunctive Datalog Queries

摘要

In this paper we present a technique for the optimization of bound queries over disjunctive deductive databases with constraints. The proposed approach consists of two distinct phases: i) the rewriting of queries for propagating bindings from the query goal into the program, and ii) the use of specialized algorithms computing rewritten queries. The rewriting of queries is based on the exploitation of a binding propagation technique which reduces the size of the data relevant to answer the query and, consequently, minimizes both the complexity of computing a single model and the whole number of models to be considered. As for general queries the rewriting technique does not ensure soundness, we present two sound and complete algorithms computing rewritten queries under brave and cautious reasoning. The efficiency of our algorithms has been proved by several experiments considering both classical search and optimization problems.