Abstract:A fundamental problem in retail business is selecting items with consideration of "cross-selling effect". Recent studies show that the problem is NP-hard. In this paper, the genetic algorithm is applied to this problem, and a quantitative analysis method for "cross-selling effect" is proposed, using quantitative association rules, and a newly developed data mining technique to identify quantitative affinities in large transaction databases. Based on the features of the genetic algorithm, a method of defining imprecise fitness function is proposed, which can be used for solving item selection problems with imprecision cross-selling effect.