Fuzzy Association Rules Discovered on Effective Reduced Database Algorithm

Fuzzy association rules can deal with continuous (numerical) attributes in database, and hence provide alternative new approach for their applications, such as supermarket basket analysis. This new approach can not only find the relations of continuous attributes, but also discrete (nominal) attributes by using crisp sets as special fuzzy sets, moreover combine them together to get good rules for analysers. However, compared with traditional models, fuzzy models generally have space and time complexity problem. We therefore develop the effective reduced database algorithm with less space and time complexity, which effectively form the transparent and knowledge based fuzzy model -reduced database table, so that we can simply discover fuzzy association rules from the reduced database table