基于敏感特征的网络钓鱼网站检测方法

Phishing detection method based on sensitive characteristics of phishing webpage

中文关键词: 网络钓鱼  敏感特征  线性分类器

英文关键词: phishing  sensitive characteristics  linear classifier

中文摘要:

      网络钓鱼(phishing)是一种在线欺诈行为,普遍存在于电子商务和电子金融中．将黑白名单方法和异常特征检测方法相结合,针对网络钓鱼网站URL异常和页面身份异常特点提出基于敏感特征的网络钓鱼网站检测方法——PhishDetector．使用黑白名单技术对URL进行拦截,对于名单中不存在的URL,提取其敏感特征,然后使用线性分类器判断该网站是否为网络钓鱼网站．实验结果表明,基于敏感特征的网络钓鱼网站检测方法,提高了网络钓鱼网站检测的正确率,显著降低了误判率．

英文摘要:

      Phishing is a kind of online fraud which is widespread in the electronic commerce and electronic banking. A method of phishing detection based on sensitive characteristics — PhishDetector is proposed. Firstly, the black/white list technology is used to intercept URL, and then the sensitive features of those URL not existing in the lists are extracted, and lastly a phishing webpage is distinguished with a linear classifier. The experimental results show that this method is better both in accuracy and false positive rate.