Research Of Association Rules Mining Based On Cluster And Matrix

From the data mining theory appeared in last 1980's, it has developed speedily. Data mining is the process of abstraction unaware, potential and useful information and knowledge from plentiful, incomplete, noisy, fuzzy and stochastic data. In this paper, we main focus on an important domain of the data mining: association rules analysis. In 1993, R.Agrawal proposed the association rules problem. After this, association rules mining obtained extensive attention of numerous experts and scholars, and it became one of the most active research directions of the data mining. This paper first introduces the concept of association rules mining and its two classical algorithms—Apriori algorithm and FP-growth algorithm an. Then according to the deficiencies of Apriori algorithm and some improved Apriori algorithm, we propose a new improved algorithm, association rules mining algorithm based on cluster and matrix—CM-Apriori algorithm. This algorithm scans the database only once, then clusters some matrixes according to the numbers of items, calculates partial matrixes to get frequent itemsets. This algorithm decreases the times of scanning of database and computation cost. Thus raises the algorithm operation efficiency effectively. The example analysis and performance study indicate that this algorithm surpasses the similar algorithm.