eCite Digital Repository

Association rule mining for both frequent and infrequent items using particle swarm optimization algorithm


Kabir, MMJ and Xu, S and Kang, BH and Zhao, Z, Association rule mining for both frequent and infrequent items using particle swarm optimization algorithm, International Journal on Computer Science and Engineering, 6, (7) pp. 221-231. ISSN 0975-3397 (2014) [Refereed Article]

Copyright Statement

Copyright 2014 IJCSE

Official URL:


In data mining research, generating frequent items from large databases is one of the important issues and the key factor for implementing association rule mining tasks. Mining infrequent items such as relationships among rare but expensive products is another demanding issue which have been shown in some recent studies. Therefore this study considers user assigned threshold values as a constraint which helps users mine those rules which are more interesting for them. In addition, in real world users may prefer to know relationships among frequent items along with infrequent ones.

The particle swarm optimization algorithm is an important heuristic technique in recent years and this study uses this technique to mine association rules effectively. If this technique considers user defined threshold values, interesting association rules can be generated more efficiently. Therefore this study proposes a novel approach which includes using particle swarm optimization algorithm to mine association rules from databases. Our implementation of the search strategy includes bitmap representation of nodes in a lexicographic tree and from superset-subset relationship of the nodes it classifies frequent items along with infrequent itemsets. In addition, this approach avoids extra calculation overhead for generating frequent pattern trees and handling large memory which store the support values of candidate item sets.

Our experimental results show that this approach efficiently mines association rules. It accesses a database to calculate a support value for fewer numbers of nodes to find frequent itemsets and from that it generates association rules, which dramatically reduces search time. The main aim of this proposed algorithm is to show how heuristic method works on real databases to find all the interesting association rules in an efficient way.

Item Details

Item Type:Refereed Article
Keywords:particle swarm optimization, data mining, genetic algorithm, frequentitemsets, lexicographic tree
Research Division:Information and Computing Sciences
Research Group:Machine learning
Research Field:Neural networks
Objective Division:Information and Communication Services
Objective Group:Information systems, technologies and services
Objective Field:Information systems, technologies and services not elsewhere classified
UTAS Author:Kabir, MMJ (Mr Mir Kabir)
UTAS Author:Xu, S (Dr Shuxiang Xu)
UTAS Author:Kang, BH (Professor Byeong Kang)
UTAS Author:Zhao, Z (Dr Zongyuan Zhao)
ID Code:93776
Year Published:2014
Deposited By:Information and Communication Technology
Deposited On:2014-08-15
Last Modified:2017-11-13

Repository Staff Only: item control page