摘要
随着企业信息化程度的提高和互联网的普及,每天都会产生海量的实时数据,而数据流挖掘则为分析海量数据提供了一种新途径。数据流挖掘中的聚类、分类、离群点检测等算法的研究取得了进展,为在持续审计中应用数据流挖掘提供了可行性。本文提出的一种基于数据流挖掘的持续审计模型,克服了传统持续审计模型对审计端的存储能力要求高、占用大量硬件资源、联机分析时间长、对异常数据的发现滞后等缺点。
With the development of enterprise informatization and the popularity of the Internet, massive real-time data are being produced every day. Data stream mining provides one novel approach to analyzing massive real-time data. In this paper the sate-of-art in this field is presented, and its availability to continuous audit is discussed. Finally, based on data stream mining, one continuous audit model is proposed, which overcomes the disadvantages of huge storage capacity requirements, long-time online analysis and the delayed finding of abnormal data.
出处
《南京审计学院学报》
2011年第1期36-40,共5页
journal of nanjing audit university
基金
国家自然科学基金(70971067/G0112)
国家社会科学基金(10BGL016)
江苏省高校自然科学研究项目(09KJD520006)
关键词
数据流挖掘
持续审计
审计模型
聚类
分类
离群点检测
data stream mining
continuous audit
auditing model
clustering
classification
outlier detection