A new fast algorithms for mining association rules in large databases

Proposes an algorithm for mining association rules in large databases. We introduce the problem of mining a large collection of basket data type transactions for association rules between sets of items with some minimum specified confidence, and presents an efficient algorithm for this purpose. The contribution of this project is threefold: (1) efficient generation for large itemsets by hash method (2) effective reduction on itemsets scan required by the division approach and (3) the option of reducing the number of database scans required Our proposed hash and division-based techniques, HD algorithm, is very efficient for the generation of candidate large itemsets where the number of candidate large itemsets generated by HD is, smaller than that by many methods such as the Apriori algorithm, DHP algorithm and DIC algorithm According to our simulation results, the proposed approach is more efficient than any existing algorithms.