Parallel Apriori Algorithm Based on the Thread Pool
Yue Shen, Zhongqian Fu, Lei Zhang, Jianyu Wang · 2012
Discovery of association rules has a significant meaning in data mining, in which the most influential algorithm is Apriori. Due to the huge size of database and increasing amount of computation, a number of parallel algorithms have already been proposed. In order to achieve high-performance parallel computing, it is necessary to reduce redundant computation and avoid too much communication between parallel tasks, and achieve load balancing. Parallel algorithms, based on multi-processor, are concerned about the distribution of parallel tasks to improve the utilization of computing resources. These methods often require a compromise in computation, communication, load balancing, etc. For the most effective control thread, we first introduce the thread pool into the Apriori parallel algorithms. This method does not require redundant communication or computation, but can achieve load balancing so as to fully utilize the computing resources.