Parallelization of the Apriori Algorithm for the Search of Frequent Elements

Abstract

There is a wide variety of techniques that increase application performance by alleviating one or more of the most important problems with today's processors. In this work, the execution time, speedup and efficiency of the linear Apriori algorithm are shown as well as parallel with the use of OpenMP. By identifying the frequent elements of transactional databases, in processing 5 thousand records the time improves in 42,078 seconds of the algorithm with openMP compared to the sequential algorithm, in the execution 8 processor cores were used.