Title: Advanced Matrix Algorithm (AMA): reducing number of scans for association rule generation

Authors: Marghny H. Mohamed, Mohammed M. Darwieesh, A.B.M. Shawkat Ali

Addresses: Faculty of Computers and Information, Assiut University, Assiut, Egypt. ' Faculty of Science, Mathematics Department, Assiut University, Assiut, Egypt. ' School of Information and Communication Technology, CQUniversity, Australia

Abstract: Existing Association Rules Mining (ARM) algorithms basically use multiple scans to extract a rule from a transaction database. Sometime ARM algorithms exit without a rule in the desktop environment due to the high volume of transactions. Matrix Algorithm (MA) is proposed to minimise this issue. However, it is a computational expensive solution. In this paper, we propose Advanced Matrix Algorithm (AMA), to generate an efficient rule by a single scan using the Boolean matrix concept. AMA is comparatively effective and efficient than traditional approaches in terms of computational cost for database scan and frequently candidate sets generation.

Keywords: ARM; association rule mining; transaction databases; AMA; advanced matrix algorithm; database scanning; association rules.

DOI: 10.1504/IJBIDM.2011.039412

International Journal of Business Intelligence and Data Mining, 2011 Vol.6 No.2, pp.202 - 214

Published online: 22 Apr 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article