摘要
针对传统关联规则Apriori算法难以适应大数据的问题,为提高可信计算平台日志数据分析效率,提出了一种基于Hadoop的可信计算平台日志分析模型。构建了日志分析模型总体框架,对非结构化原始日志数据进行垂直划分,采用分布式文件存储系统,结合MapReduce编程模式给出一种分布式Apriori并行垂直算法。通过日志挖掘建立用户行为关联规则库,并采用规则匹配实现对用户异常行为的检测。理论分析和实验数据证明,该模型在大数据环境下能够有效提高日志分析效率。
To solve the problem that the Aprioro algorithm can not adapt to mass data,and to improve the efficiency of analyzing log data of Trusted Computing Platform,a trusted computing log analysis model was constructed based on Hadoop platform.The framework of log analysis model was constructed.A parallel vertical algorithm of Apriori was proposed in MapReduce programming model and distributed file storage system by dividing the original non-structured log data into vertical partitioning pattern.Through log mining,the user behavior association rule base is set up.By rule matching,the user＇s abnormal behaviors are is detected.Mathematic analysis and simulation experiments show that the model can effectively improve the efficiency of log analysis in mass data environment.
作者
严嘉维
张琛
李成蹊
孙权荣
张程誉
谭玉霖
YAN Jia -wei 1,ZHANG Chen 2,LI Cheng-xi 3,SUN Quan-rong 2,ZHANG Cheng-yu 4,TAN Yu-lin 2(1.School od Art, Hubei University of Technology, Wuhan 430020,2.Department of Information Security, Naval University of Engineering, Wuhan 430033;3.Office of China International Futures Co.,Ltd. Wuhan 430026, China, Wuhan 430026;4.Cadet Brigade, Naval University of Engineering, Wuhan 430033, Chin)
出处
《软件导刊》
2018年第8期71-75,共5页
Software Guide
