Mining association rules with frequent closed itemsets lattice

Abstract

One of the most important tasks in the field of data mining is the problem of finding association rules. In the past few years, frequent closed itemsets mining has been introduced. It is a condensed representation of the data and generates a small set of rules without information loss. In this paper, based on the theory of Galois Connection, we introduce a new framework called frequent closed itemsets lattice. Compared with the traditional itemsets lattice, it is simple and only contains the itemsets that can be used to generate association rules. Using this framework, we get the support of frequent itemsets and mine association rules directly. We also extend it to fuzzy frequent closed itemsets lattice, which is more efficient at the expense of precision.