Using association rules to add or eliminate query constraints automatically

Abstract

Much interesting work has been done on the use of semantic associations for optimizing query execution. Our objective is to study the use of association rules to add or eliminate constraints in the where clause of a select query. In particular, we take advantage of the following heuristics presented by Siegel et al. (1992): i) if a selection on attribute A is implied by another selection condition on attribute B and A is not an index attribute, then the selection on A can be removed from the query; ii) if a relation R in the query has a restricted attribute A and an unrestricted cluster index attribute B, then look for a rule where the restriction on A implies a restriction on B. The contribution of our work is twofold. First, we present detailed algorithms that apply these heuristics. Hence, our ideas are easy to implement. Second we discuss conditions under which it is worth applying these optimization techniques, and we show the extent to which they speed up query execution.