大数据环境下网络入侵检测技术的探究  

Research on Network Intrusion Detection Technology in Big Data Environment

作　　者：蓝文奇 LAN Wen-qi(Guangdong Guangle Expressway Co.,Ltd.,Guangzhou Guangdong 511533)

机构地区：[1]广东广乐高速公路有限公司

出　　处：《数字技术与应用》2019年第5期197-199,共3页Digital Technology & Application

摘　　要：互联网为信息资源的交流提供了高效而便捷的全新方式,但同时信息资源的安全性也更加重要,信息资源的盗用、入侵、甚至毁坏给互联网的信息资源带来了严重安全威胁。而作为动态安全系统最核心的技术之一,入侵检测技术在网络防御体系中起着极为重要的作用,它是静态防护转化为动态防护的关键,也是强制执行安全策略的有力工具。本文将在阐述网络入侵检测常见方法的基础上,从数据流角度,提出动态的集成PU学习数据流分类的入侵检测方法,在验证数据集上进行突变漂移和逐渐漂移的比较实验表明,该方法具有较好的分类性能。The Internet provides an efficient and convenient new way for the exchange of information resources,but at the same time the security of information resources is even more important.The theft, invasion,and even destruction of information resources of the Internet have brought serious security threats.As one of the most core technologies of dynamic security system,intrusion detection technology plays an extremely important role in network defense system.It is the key of static protection conversion to dynamic protection, and it is also a powerful tool to enforce security strategy.Based on the description of common network intrusion detection methods,this paper proposes a dynamic intrusion detection method that integrates PU learning data stream classification from the data flow point of view.The method has good classification performance.

关 键 词：大数据 网络 入侵检测 PU学习算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]