An Improved DCI_Closed Algorithm

Abstract

Frequent itemsets mining will not conducive to data analysis. Because of frequent itemsets mining produce very large amount of frequent itemsets in the case large data. Frequent closed itemsets provides a lossless frequent item sets, the smallest representation. This paper aims at the problem of shortage of the DCI_Closed algorithm in the process of mining efficiency, puts forward a kind of improved DCI_Closed algorithm to improve the efficiency of mining, the algorithm references the concept of co-occurrence itemsets and use the nature of the co-occurrence itemsets to prune operation on the 1-itemsets. And improve the efficiency of the algorithm. The experimental results show that the improved DCI_Closed algorithm is better than DCI_Closed algorithm on running time.