Title: An efficient algorithm for mining frequent closed itemsets in dynamic transaction databases

Authors: Ruili Wang, Luofeng Xu, Stephen Marsland, Ramesh Rayudu

Addresses: Institute of Information Sciences and Technology, College of Sciences, Massey University, New Zealand. ' Institute of Information Sciences and Technology, College of Sciences, Massey University, New Zealand. ' Institute of Information Sciences and Technology, College of Sciences, Massey University, New Zealand. ' Institute of Information Sciences and Technology, College of Sciences, Massey University, New Zealand

Abstract: In this paper we propose an extension algorithm to CLOSET+, one of the most efficient algorithms for mining frequent closed itemsets in static transaction databases, to allow it to mine frequent closed itemsets in dynamic transaction databases. In a dynamic transaction database, transactions may be added, deleted and modified with time. Based on two variant tree structures, our algorithm retains the previous mined frequent closed itemsets and updates them by considering the changes in the transaction databases only. Hence, the frequent closed itemsets in the current transaction database can be obtained without rescanning the entire changed transaction database. The performance of the proposed algorithm is compared with CLOSET+, showing performance improvements for dynamic transaction databases compared to using mining algorithms designed for static transaction databases.

Keywords: data mining; frequent closed itemsets; dynamic transaction databases.

DOI: 10.1504/IJISTA.2008.017275

International Journal of Intelligent Systems Technologies and Applications, 2008 Vol.4 No.3/4, pp.313 - 326

Available online: 22 Feb 2008 *

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