fuzzy association analysis, fuzzy quantifiers, confidence, four-fold table, confirmation measures

In this contribution, we propose a generalization of a four-fold table known from classical statistics to the case when fuzzy sets represent attributes in a data set. We define fuzzy quantifiers on the generalized four-fold table and investigate various classes of such quantifiers. We also study properties of fuzzy four-fold tables and several quality measures based on them and relations between the choice of used t-norms and chosen quality measures. We do not only present a new theoretical approach to fuzzy association analysis but, based on our results, we conclude this paper with a short discussion on the best choice of t-norm used for mining fuzzy associations.