A Novel Spark-Based Algorithm for Mining Frequent Utility Patterns

We are now living in the era of big data thanks to technological innovation. It is very important to discover valuable information in massive data. High utility itemset mining and frequent itemset mining are commonly used techniques in data mining, but they only consider a single factor, ignoring some frequent and efficient itemsets. So as to mine more valuable information, a method for mining skyline frequent utility patterns is proposed. Facing multi-dimensional and large-scale datasets, how to quickly find skyline points is a new challenge. In this article, a parallel algorithm is proposed within the framework of Spark, using Transformations and Actions operators to mine skyline patterns. Through the analysis of experimental data, it is proved that the proposed method has high computing efficiency and high search efficiency, and can effectively mine skyline frequent utility itemsets.