基于数据挖掘的Web服务器攻击检测

Detection of Web Server Attack Based on Data Mining

随着Web技术的飞速发展,越来越多的单位使用Web方式为用户提供服务,黑客的矛头也逐渐转向了Web服务器。文中研究了如何使用数据挖掘的方法对Web服务器日志进行安全性分析,从而快速发现攻击,并针对Web服务器日志特有的格式,对现有的基于相似系数的异常点检测算法进行了改进,使其符合研究的要求,同时构建了两个安全性分析的数据挖掘模型。通过对现有Web日志记录的挖掘,验证了所述方法的可行性。

With the development of Web technology, more and more services are provided. Web Server has become a main target of hackers. The paper discusses how to analyze Web server log for detecting attack by using data mining method. According to the special format of Web log, this paper provides an improved outlier detection method based on similar coefficient. In order to finish this task, two data mining models are made. The test results are also given, thus to verify their feasibility.