Using rule-free sets to find frequent itemsets

Given a large set of data, extracting frequent itemsets in this set is a challenging job in data mining. Many efficient algorithms have been proposed in the literature. The idea presented in this paper is to extract a condensed representation of the frequent itemsets called rule-free sets, instead of extracting the whole frequent itemsets collection. We show that this condensed representation can be applied to regenerate all frequent itemsets and their exact frequencies without any access to the original data. An algorithm, RFS-MINER, is presented to extract the frequent rule-free sets and practical experiments show that this representation can be extracted very efficiently. We compared it with another representation in the literature called frequent closed sets and in nearly all experiments we have run, the rule free sets have been extracted much more efficiently than frequently closed sets.