基于行为分析的钓鱼页面检测技术研究  

Technique of Detection Phishing Page Detection based on Behavior Analysis

作　　者：赵功勋 陈运[1] 杨义先[2] 

机构地区：[1]成都信息工程学院计算机学院,四川成都610225 [2]北京邮电大学,北京100876

出　　处：《成都信息工程学院学报》2014年第5期464-470,共7页Journal of Chengdu University of Information Technology

基　　金：国家自然科学基金资助项目(60873216);四川省科技支撑计划重点资助项目(2012GZ0017)

摘　　要：针对通过黑白名单匹配的检测方法检测钓鱼页面准确率低的特点,提出基于钓鱼者行为的图状链接结构特征,对钓鱼页面进行分析,引入数据挖掘的频繁子图挖掘算法,对数据库中积累的数万个钓鱼页面进行子图模式挖掘,提取钓鱼页面的共同子图结构特征,检测网络钓鱼行为。实验结果发现,在加入了子图特征的钓鱼页面检测方法中,检出率能达到80%.因此,基于行为的钓鱼页面检测提高了钓鱼页面检测的能力,并且挖掘出的子图模式为钓鱼者的行为提供了依据。For the accuracy of testing phfishing pages matched by the black and white lists are relatively low,.So in this paper,we putPutting forward to analysisanalyse of the link structure of anglers in phfishing pages based on the graph-like behavior,introduceintroducing data mining algorithm for mining frequent sub-graphs,minemining tens of thousands of phishing pages accumulated in the database based on sub-graphs pattern,extractextracting the structural features of the commonsame sub-graphs in the phfishing page.structural features,Andand then thedetecting behaviors of phishing pages are detected.Experiment results show that Experimentally found that detection of phfishing page based on the subgraph feature in the phfishing pages can approach to eighty percents.So,detecting phishing pages detection based on the behavior analysis can improve the ability to detect phishing pages and the excavated sub-graphs model provides a basis for anglers behavior.

关 键 词：信息安全 网络安全 钓鱼 行为分析 数据挖掘 频繁子图 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]