摘要
针对有时间约束的泛化序列模式的挖掘问题 ,提出了一种有效的挖掘方法 .与已有的算法相比 ,主要通过采取两种技术来提高效率 ,一是事先找出每个数据序列支持的序列模式 ,从而去除了时间因素 ,用一个快速算法来解决匹配问题 ;二是在数据序列重复较多时采用直接求交的方法 .
An efficient algorithm for mining generalized sequence patterns is presented. The algorithm employs two ways for improving the efficiency. One is to find in advance the sequential patterns supported by each data sequence, thereby the time constraints are eliminated and a fast matching algorithm can be applied. The other is to compute the intersection of sequential patterns. Finally, a mining algorithm is given based on partitioning the database.
出处
《浙江大学学报(理学版)》
CAS
CSCD
2002年第4期415-422,共8页
Journal of Zhejiang University(Science Edition)