A New Approach for Frequent Pattern Mining

In this paper, we review the partition algorithm proposed for mining frequent itemsets and we propose AdvPartition, a new algorithm, which introduces several improvements to the classic Partition algorithm. Our goal was the optimization of the most time consuming phase of Partition algorithm i.e. the Database Scans. In a thorough experimental evaluation of our algorithm on standard benchmark data from the literature, our algorithm outperforms previous work upto an order of magnitude.