A novel parallel frequent itemset mining algorithm for automatic enterprise.

Heterogeneity, volume and real-time velocity of manufacturing data affect the business efficiency within the process for analyzing data in Robotic Process Automation (RPA). A novel parallel frequent itemset mining algorithm based on MapReduce (PMRARIM-IEG) is designed to improve the business efficiency. The algorithm is designed to address issues such as the CanTree's excessive space usage, the inability to dynamically set the support threshold, and the time-consuming data transmission during the Map and Reduce phases. Experiments show that the proposed algorithm has lower memory usage and higher parallel efficiency than the traditional parallel frequent itemset mining algorithm. [ABSTRACT FROM AUTHOR]