Improving Association Rules Mining by Hashing Algorithm

Abstract

Association rule mining, is one of the most important and well researched techniques of data mining. It is the process of discovering large itemsets appeared in a sufficient number of transactions. Large itemsets from a huge number of candidate large itemsets are dominating factor for the overall data mining performance. This paper presents mining association rules among items in a large database of sales transactions. To address this problem, it present an effective hash-based algorithm for the candidate set generation. This system applies an algorithm DHP (Direct Hashing and Pruning) on application cosmetic sales data to generate frequent association patterns. Generation of smaller candidate sets enables to effectively trim the transaction database size at a much earlier stage of the iterations, thereby reducing the computational cost of later iterations significantly. The experimental results of our system are also discussed in this paper.