摘要
本文通过在传统序列模式中加入了多序列限制,提出了多序列模式的概念。在Prefixspan算法的基础上,增加基于多序列约束的剪枝技术,实现了多序列环境下的多序列模式挖掘算法(MS_Prefixspan)。实验结果表明,该算法的时间性能要优于Prefixspan算法,并且能够有效减少模式的数量,使得挖掘结果更具目标性。
This paper put forward the concept of multiple sequential patterns by adding multi-sequence constraint on the traditional sequential patterns. Based on the Prefixspan algorithm we designed multiple sequential patterns mining algorithm (MS_Prefixspan). The MS_Prefixspan algorithm use pruning technique of multiple sequence constraint. The experimental results show that our proposed algorithm outperforms Prefixspan,and can reduce the number of patterns,making mining results more targeted.
出处
《微计算机信息》
2010年第36期195-196,58,共3页
Control & Automation
关键词
数据挖掘
序列模式挖掘
多序列模式
Data Mining
Sequential patterns mining
Multiple Sequential patterns