摘要： 给出单模式、二模式和三模式3种序列模式发现的基本概念，给出二模式和三模式的表示方法。该表示方法不会产生实际不存在的候选序列，从而有效地缩小候选空间，提高序列模式的计算速度。结合Apriori方法，基于3种基本模式，应用无冗余的模式增长原则和三级动态优化方法，提出一种序列模式发现的结构化动态优化方法。

关键词:

Apriorimso-font-kerning: 0pt, mso-ascii-font-family: 'Times New Roman', mso-hansi-font-family: 'Times New Roman'">方法, mso-ascii-font-family: 'Times New Roman', mso-hansi-font-family: 'Times New Roman'">单模式, 二模式, 三模式, mso-font-kerning: 0pt, mso-ascii-font-family: 'Times New Roman', mso-hansi-font-family: 'Times New Roman'">动态优化mso-bidi-font-family: 宋体">

Abstract: This paper proposes the basic concepts of single item schema, two item schema and three or more item schema, and presents the representation method of two item schema and three or more item schema, by which candidate sequences not exist in fact are not produced, so that the number of candidates is reduced and the speed of computing sequence schema is enhanced. Based on Apriori method and three schemas, by using irredundant produce schemas principle and three-level optimum method, a structural dynamic optimum method of sequential pattern discovery is proposed.

Key words: Apriori method, single item schema, two item schema, three or more item schema, dynamic optimum

中图分类号: 

• TP311.13