Semantic Optimization of Preference Queries

TL;DR

This paper introduces semantic optimization techniques for preference queries in databases, utilizing integrity constraints to improve efficiency and eliminate redundancies in winnow operations.

Contribution

It presents novel semantic optimization methods for preference queries that leverage integrity constraints to enhance query efficiency and reduce redundancy.

Findings

Redundant winnow operators can be identified and removed.

A more efficient algorithm for computing winnow is proposed.

Conditions for applying optimization techniques are characterized as constraint satisfiability problems.

Abstract

The notion of preference is becoming more and more ubiquitous in present-day information systems. Preferences are primarily used to filter and personalize the information reaching the users of such systems. In database systems, preferences are usually captured as preference relations that are used to build preference queries. In our approach, preference queries are relational algebra or SQL queries that contain occurrences of the winnow operator ("find the most preferred tuples in a given relation"). We present here a number of semantic optimization techniques applicable to preference queries. The techniques make use of integrity constraints, and make it possible to remove redundant occurrences of the winnow operator and to apply a more efficient algorithm for the computation of winnow. We also study the propagation of integrity constraints in the result of the winnow. We have identified necessary and sufficient conditions for the applicability of our techniques, and formulated those conditions as constraint satisfiability problems.