基于IRST的并行时序模式挖掘算法

Parallel mining sequence pattern algorithm based on IRST

提出一种建立在集群式高性能计算机上基于互关联后继树的并行时序模式挖掘算法,将数据线段化、树的建立及模式发现在多处理机上进行并行处理,有效地改进了算法的执行效率。实验结果表明,此算法较之串行算法有较高的效率。

This paper proposed a parallel mining sequence pattern algorithm based on IRST. It distributeed mining tasks to multiprocessor, including sequence segment, create SIRST and finding frequent patterns. Experiments show that it is more efficient.