A Hash based Mining Algorithm for Maximal Frequent Item Sets using Hashing

Abstract: Data mining plays important role in many applications like market-basket analysis, cross marketing field etc. In data mining, Association Rule Mining (ARM) finds the interesting relationship between across of various items in a given database. In this paper, we propose a new association rule mining algorithm called Hash Based Frequent item sets-Quadratic probing (HBFI-QP) in which hashing technology is used to resolve primary collisions in vertical data format of the data base. But Quadratic probing also suffer from secondary clustering. This secondary clustering problem solve by using double hashing technique (HBFI-DH). The proposed technique generates the exact set of maximal frequent item sets directly by removing all non-maximal item sets. The proposed technique access the data fastly and efficiently compare with other hashing technique.

Keywords: Frequent Item Sets, data mining, Association Rule Mining (ARM), double hashing technique (HBFI-DH).