摘要
攻击者常以政府,校园网站为目标进行网页篡改,来达到抹黑国家形象,发展黑产等目的。现有对于网页篡改的检测研究主要以网站结构检测或者历史对比为主,存在误报率高的问题。该系统针对网页篡改,设计一个以网页页面为对象,从敏感词,违规图片特征,流量-时间特征三方面进行识别,并由基于决策树算法的分析器进行网页篡改判断的系统。最后进行了实验验证,在保证不同场景下识别篡改成功率均超过92%的同时,控制误报率未超过2.8%的检测目的。
Attackers often use the government and campus websites as the target to tamper with the web to achieve the goal of discrediting the country’s image and developing black products.The existing research on the detection of web tampering mainly focuses on website structure detection or historical comparison,and there is a problem of high false positive rate.This system aims at the tampering of web pages,and designs a system that uses web pages as objects to identify sensitive words,illegal picture features,and traffic-time characteristics,and uses a analyzer based on decision tree algorithm to detect web tampering judgments.And at the last of this paper,it was proved with an experiment that the success rate of tampering detection exceeded 92%in different scenarios,while the false detection rate did not exceed 2.8%.
作者
李柯言
刘晓东
LI Ke-yan;LIU Xiao-dong(Wuhan Research Institude of Posts and Telecommunications,Wuhan 430070,China;Wuhan Hongxu Information Technology Co.,Ltd.,Wuhan 430070,China)
出处
《电子设计工程》
2020年第18期16-19,24,共5页
Electronic Design Engineering
关键词
网页篡改
图像识别
决策树
低误报率
Web page tampering detection
image detetion
decision tree
low false detection rate
作者简介
李柯言(1996-),男,辽宁开原人,硕士。研究方向:网络安全。
