【摘要】 序贯模式是数据挖掘领域的重要研究课题之一。鉴于目前国内外在此方面的研究主要是集中在对交易数据的分析处理上,而从数据挖掘本身的角度考虑,交易数据库与常用的关系数据库两者之间却存在着很大的差异,这就要求新的挖掘算法与之相适应。本文针对如何从一般关系数据库中进行序贯模式的知识发现问题做了初步的研究和探讨,并在此基础上形成了一种基于时间窗口的序贯模式挖掘算法TW_SP。更多还原

【Abstract】 Mining sequential patterns is an important topic in data mining research. At present, the researches and applications are mainly focused on analyzing transaction data. With respect to data mining, the transaction database is very different from the general relational database. In this paper we make relevant study and discussion on how to mine sequential patterns from general relational database. Then, a time-window based algorithm for mining sequential patterns (TW SP) is proposed.更多还原