Discovery of association rules is an important problem in the area of data mining. For this problem, how to efficiently count large itemsets is the major work, where a large itemset is a set of items appearing in a sufficient number of transactions. Most of previous study generated candidates and tested them to gain the large itemsets. It is very time-exhausted. Therefore, in this paper, we propose the DPC-I algorithm, which Directly Prunes non-large itemsets and Counts the others, to find a large itemset of a certain Interesting size. In the DPC-I algorithm, given a k, we efficiently construct L_k based on L_2, instead of step by step, where L_k denotes the set of large k-itemsets with minimum support. We conduct several experiments using different synthetic transaction databases. The simulation results show that the DPC-I algorithm outperforms the Apriori algorithm and FP-Growth algorithm in the execution time for all transaction database settings.
展开▼