An Adjacency Matrix Based Apriori Algorithm for Frequent Itemsets Mining

Abstract

Finding frequent itemsets is a most researched field in data mining. Currently, the finding of frequent itemsets problem’s solution has been proposed by many researchers. The Apriori algorithm is the basic algorithm for frequent itemsets mining. In Apriori algorithm, there are main two issues: scanning the database multiple times and generating a large number of candidate sets. In recent years several improved apriori algorithms have been defined and evaluated to improve efficiency. Our main goal is to define a new optimized algorithm and to compare its performance with the existing algorithms. The main focus of our work is to propose a new optimized algorithm and to compare its performance with the state of the art methods. In proposed work, adjacency matrix will be employed in order to improve the operating efficiency and eliminate the candidate sets. In a proposed system not require the pruning step. Performance of the proposed method will be evaluated on existing datasets. A secondary data set is used to find frequent itemsets with using our proposed algorithm and existing algorithm. The effect of our proposed algorithm is presented.