An efficient approach to discovering frequent patterns from data cube using aggregation and directed graph

Abstract

In this paper, an algorithm has been proposed for mining frequent itemsets from data cube. Discovering frequent itemsets has been a key process in association rule mining. The major drawbacks of traditional algorithms are that lot of time consumed to find candidate itemsets and lot of memory to store them. Proposed algorithm discovers frequent itemsets using aggregation function and directed graph. It saves lot of memory consumption in candidate generation. It uses aggregation function for dimension reduction and directed graph for candidate itemsets generations. Experimental results show that the proposed algorithm can quickly discover candidate itemsets and effectively mine potential frequent patterns. © 2015 ACM.