A Modified Apriori Algorithm for Fast and Accurate Generation of Frequent Item Sets

The Classical Apriori Algorithm (CAA), which is used for finding frequent item sets in Association Rule Mining, consists of two main steps; the join step for generating candidate item sets and the prune step for eliminating candidate item sets that are not frequent. The CAA despite its simplicity has some limitations; the generation of a large number of candidate item sets, the generation of many combinations that never occur in the database as well as the need to perform several full database scans when generating frequent item sets. In this research, a Modified Apriori Algorithm (MAA) is proposed to address the problem of generating many combinations that never occur in the database by using a row-wise combination generation technique. A comparison of the results of the proposed algorithm against the Classical Apriori Algorithm shows that the proposed algorithm is faster and more efficient. The MAA was implemented on transaction databases and the results were compared against results from four (4) other Improved Apriori Algorithms for efficiency. The results of the comparative analysis showed that the MAA was more efficient in terms of execution time than the other Improved Apriori Algorithm.