摘要
基于序列模式图可以定义某些结构化的新知识,如分支模式、重复模式以及它们的组合———结构模式.在基于事件的数据或序列数据中发现结构模式是后序列模式挖掘的任务.基于Apriori算法思想,介绍结构模式的主要形式———并发分支模式及其挖掘算法,并通过示例解释挖掘算法的实现过程.该算法的基本思想是在粗并发分支模式的基础上生成并发分支模式,在并发分支模式生成过程中采用自底向上的方法.基于Apriori的并发分支模式挖掘算法的提出为进一步挖掘结构模式提供了重要的理论支持.
On the basis of the Sequential Patterm Graph（SPG）, some new structural knowledge such as branch pattern,iterative pattern and structural pattern were defined. Mining structural pattern in event based data or sequence data is the task of Post Sequential Patterns Mining （PSPM）. Based on Apriori property, in this paper, we concentrate on concurrent branch patterns and its mining algorithm. A con- crete example is also given to illustrate the algorithm. The basic idea of the mining algorithm is to generate concurrent branch pattern from rough concurrent branch pattern. Bottom-up method is adopted in the process of pattern generating. The Apriori-based concurrent branch pattern mining algorithm provided important theoretical support for further mining structural pattern.
出处
《沈阳化工学院学报》
2006年第1期59-62,共4页
Journal of Shenyang Institute of Chemical Technolgy
基金
辽宁省教育厅科学研究计划资助项目(20040287)
关键词
后序列模式挖掘
并发分支模式
序列模式图
post sequential patterns mining
concurrent branch patterns
sequential patterns graph
