Fuzzy Logic -based Pre-processing for Fuzzy Association Rule Mining

Abstract

Conventional Association Rule Mining (ARM) algorithms usually deal with datasets with categorical values and expect any numerical values to be converted to categorical ones using ranges (Age = 25 to 60). Fuzzy logic is used to convert quantitative values of attributes to categorical ones so as to eliminate any loss of information arising due to sharp partitioning (using ranges) and then generate fuzzy association rules. But before any fuzzy ARM algorithm can be applied, there is substantial amount of fuzzy-logic-based pre-processing that needs to be done on the dataset. This paper describes, in detail, a methodology to do this pre-processing which first involves using fuzzy clustering to generate fuzzy partitions and then use these partitions to get a fuzzy version (with fuzzy records) of the original dataset. Ultimately, the fuzzy data (fuzzy records) are represented in a standard manner such that they can be used as input to any kind of fuzzy ARM algorithm, irrespective of how it works and processes the fuzzy data. And finally, we also show how existing algorithms, like Apriori, ARMOR, and FPGrowth can be modified, especially in the manner in which they count itemsets, to mine data in a fuzzy environment.