Scalable Query Reformulation Using Views in the Presence of Functional Dependencies

Abstract

The problem of answering queries using views in data integration has recently received considerable attention. A number of algorithms, such as the bucket algorithm, the SVB algorithm, the MiniCon algorithm, and the inverse rules algorithm, have been proposed. However, integrity constraints, such as functional dependencies, have not been considered in these algorithms. Some efforts have been made in some inverse rule-based algorithms in the presence of functional dependencies. In this paper, we extend the bucket-based algorithms to handle query rewritings using views in the presence of functional dependencies. We build relationships between views containing no subgoal of a given query and the query itself. We present an algorithm which is scalable compared to the inverse rule-based algorithms. The problem of missing query rewritings in the presence of functional dependencies that occurs in the previous bucket-based algorithms is avoided. We prove that the query rewritings generated by our algorithm are maximally-contained rewritings relative to functional dependencies.