Some algorithms for evaluating fuzzy relational queries

Abstract

An important issue in extending DBMS functionnalities is to allow the expression and execution of flexible queries in order to make these systems able to satisfy user needs more closely. To deal with this problem it is necessary to define new evaluation methods since standard (crisp) algorithms and access paths have proved to be inappropriate. A challenge in solving this problem is to keep the additional costs entailed by these new querying capabilities at a reasonable level. For this purpose we propose a set of query evaluation algorithms based on techniques expected on the one hand to restrict the amount of tuples to be transferred from disk to main memory and on the other hand to limit the computations which apply to these tuples. This work only concerns the evaluation of basic extended relational operators and may be seen as a first step towards the design of a fuzzy query optimizor. For each of the proposed methods, we suggest some useful measures for determining to what extent the corresponding algorithms are acceptable. Finally, we present an experimental framework for measuring the performance of these algorithms.