About fuzzy query processing-the example of the division

This paper describes a general method for the processing of complex fuzzy queries addressed to usual databases. The proposed approach relies on a low-level interface to a regular DBMS, whose purpose is to derive an evaluation algorithm involving some SQL queries, some functions aiming at the computation of the grades, as well as some heuristics intended to optimize the number of data accesses. We show how this method can be applied to the processing of an extended division and, in this context, we give some experimental results concerning the extra cost brought by the treatment of graduality.