A survey on frequent pattern mining: Current status and challenging issues

Abstract

Discovering association rules in huge databases is a core topic of data mining. This survey study aims at giving an overview of the previous researches done in this field, evaluating the current status of the work done and envisioning gaps in the current knowledge. The problem of mining association rules can be generalized in to two steps: (1) Finding all frequent itemsets and (2) generating rules from these itemsets. The first sub-task, which is to determine the frequent itemsets, is computationally expensive process. Counting the occurrences of itemsets requires a considerable amount of processing time. As a consequence, number of algorithms are proposed in literature for mining the frequent itemsets. Present study reviews frequent pattern mining algorithms and other related issues available in the literature. © 2010 Asian Network for Scientific Information.