Title: An efficient graph-based approach to mining association rules for large databases

Authors: Lee-Wen Huang, Ye-In Chang

Addresses: Department of Computer Science and Engineering, National Sun Yat-Sen University, No. 70, Lienhai Rd., Kaohsiung 80424, Taiwan. ' Department of Computer Science and Engineering, National Sun Yat-Sen University, No. 70, Lienhai Rd., Kaohsiung 80424, Taiwan

Abstract: The task of data mining is to find the useful information within the incredible sets of data. One of important research areas of data mining is mining association rules. If we can find these relations by mining association rules, we can provide better selling strategy to gain more customers| attentions. However, in some applications, the large itemsets may not always correlate with each other. In this paper, we propose a new graph-based algorithm to discover the association rules. It represents the large itemsets as a graph, which constructs a graph based on L2. Then, by dividing the items to several groups, the association rule can be mined efficiently. We conduct several experiments using different synthetic transaction databases. The simulation results show that the GAR algorithm outperforms the FP-growth algorithm in the execution time for all transaction databases.

Keywords: association rules; data mining; graph-based mining; large itemsets; knowledge discovery.

DOI: 10.1504/IJIIDS.2009.027686

International Journal of Intelligent Information and Database Systems, 2009 Vol.3 No.3, pp.259 - 274

Published online: 07 Aug 2009 *

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