Fuzzy Association Rules: a Two-Sided Approach
Martine De Cock, Chris Cornelis, Etienne E. Kerre · 2003
We bring to the surface the fundamental two-sidedness of knowledge in the framework of association rules, until now only slumberingly present in the measures of support and confidence. We identify the set of positive as well as the set of negative examples which are not necessarily complementary. Taking this into account we introduce new quality measures comprising the existing ones. Finally, we carefully examine the generalization of our findings to fuzzy association rules.