摘要
加权关联规则挖掘是告警相关性分析的重要手段。本文引入了比例加权支持度的概念,提出了一种基于加权关联模式树的加权关联模式挖掘算法。实验表明,本算法与MINWAL(O)算法相比,时间效率有了明显提高,节约了存储空间,告警相关性分析的准确性也得到了提高。
The mining of weighted association rules is a primary method used in alarm correlation analysis. In this paper, a new algorithm, weighted association pattern mining from weighted associated pattern tree is proposed, in which a new measure, ratio-weighted support, is developed. Experimental results demonstrate compared with MINWAL（O） algorithm, this algorithm has higher temporal etticiency and requires less memory. Also, the accuracy of alarm correlation analysis is improved.
出处
《电信科学》
北大核心
2007年第11期57-60,共4页
Telecommunications Science
基金
国家自然科学基金资助项目(No.60572091)
关键词
告警相关性分析
加权关联模式
比例加权支持度
加权关联模式树
alarm correlation analysis, weighted association pattern, ratio-weighted support, weighted associated pattern tree
