The Research Of The Frequent Item Sets Discovery Algorithm Of Association Rules Data Mining

Data Mining is one of the most active research fields, especial in the fields of artificial intellegince and database reasearch. Data Mining is a kind of process that reveals potential useful knowledge from massive data. The association rule mining is a main research aspect of data mining. And the discovery of the frequent item sets is a key problem of the association rule mining.In the thesis we fully describe the most typical discovery algorithms of the frequent item sets, Apriori algorithm and Apriori_Tid algorithm, and discuss the advantages and the disadvantages of some existent improved methods. Based on these works we present several new effective improved methods for the discovery algorithm of the frequent item sets. First, the dynamic self-adaptive method is presented in order to obviously reduce the times of scanning transactions database. Second, we present the method of gradually reducing the length of transaction records to consumedly improve the efficiency of the frequent item sets discovery algrothm. At last, on the aspect of searching the item sets, according to the orderal nature of the item sets, the binary searching method is adopted to imporve the searching efficiency. Base on these new improved methods on the above, and fully considering the advantages of the existing improved methods, the author presents a new discovery algorithm of the frequent item sets—Apriori_Auto Algorithm. The algorithm has obvious and effective improvements on the aspects of reducing the database scanning times, reducing the candidate item sets amount and improving time efficiency of the searching.