Detecting irregular fuzzy coherent rules in a predefined taxonomy

Real world applications use quantitative data instead of binary data. Fuzzy association rule mining approaches are introduced to mine quantitative data and derive interesting features among them. Fuzzy multi-level association rule mining techniques mine significant rules at various levels of hierarchical data using interesting measures like support and confidence. Coherent rule mining is introduced to extract significant rules without using interesting measures. In this paper we introduce a new form of generalized fuzzy coherent rules called irregular fuzzy rules when mining under multi-level taxonomy. These rules lack consistency when traversed from one taxonomical level to another leading to take decisions in market basket applications.