A Method to Optimize Apriori Algorithm for Frequent Items Mining

This paper studies the fundamental problems of mining association rules. Based on the summary of classical mining algorithm and the inherent defects of Apriori algorithm, some related improvements are researched. In order to avoid scanning the database multiple times, the database mapping method is changed in this research. Meanwhile, after the support of candidate item sets is get, each candidate item set should be determined whether it is a frequent item set or not based on the prior knowledge of Apriori algorithm. If the candidate item sets generated by the element of the existing frequent item sets are certainly not frequent item sets, the element is not necessary to connect with others, which leads to an optimized connecting step. Lastly, for Apriori algorithm, the intersection operation is introduced to address the disadvantages that it takes many time costs to match with candidate item sets and transaction pattern. Through these improvement strategies, the optimized algorithm is presented and its advantages are explained in theory. And furthermore, to verify the effectiveness, the optimized algorithm has been applied to the floating car data. The experiments results show a shorter execution time and a higher efficiency under different supports and confident levels.