Elite Frequent-Utility Pattern Mining in Hadoop Environments

With the extensive research of skyline frequent-utility pattern mining in the field of data mining, various algorithms emerge in an endless stream, but because these algorithms are limited by the size of the data set, traditional pattern algorithms do not perform well in medium and large databases. Therefore, this paper proposed the EFUP_MR algorithm by adding thresholds and applied it to the hadoop platform. It can run in medium and large databases, solve the problem of low efficiency in the traditional pattern, and because of the threshold setting, it can help users obtain the required elite itemsets more accurately. Finally, through experiments in different types of data sets, the proposed algorithm performance in larger datasets is proved.