Optimization of Preference Queries with Multiple Constraints

Nowadays, the efficient integration of preference querying into standard database technology is an important issue. In some instances, preference queries challenge traditional query processing and optimization. In this paper we study preference database queries involving hard constraints over multiple attributes belonging to several relations. The main bottleneck for such queries is the computation of the cartesian product which may lead to high memory and computation costs. We develop algebraic optimization techniques to transform a preference query with hard constraints in order to enable its efficient processing by database engines. For this purpose, we show a dominance criterion and we introduce rewriting techniques to eliminate dominated tuples before building the cartesian product and therefore speed up the evaluation. These techniques lead to novel preference transformation laws and extend previous developed rules.