Mining association rules based on an improved Apriori Algorithm

In this paper, we first describe the classical Apriori Algorithm. And then, we present the defects exist in this algorithm. Such as spending a lot of time to produce the candidate item-sets, scanning the database in a simple method, and so on. At last we promote our improved Apriori Algorithm which consists of three parts. The first part is reducing the number of judgments, and the second part is reducing the number of candidate frequent item-sets. The last part is optimizing the database. And the experimental results proved our improvement.