Abstract: With the rapid development of online business and online banking, Phishing attacks are becoming increasingly rampant. This paper proposes an algorithm to detect Phishing Web page. The algorithm extracts anomaly features from Web page based on DOM model, then uses Back Propagation(BP) neural networks to detect the abnormal levels of the page, and applies linear model for page classification. The implementation of the algorithm successfully detects most Phishing pages, protecting users from Phishing attacks effectively.

Key words: Phishing, Web page detection algorithm, Back Propagation(BP) neural network

摘要： 网络钓鱼(Phishing)攻击在电子商务和电子金融中普遍存在。该文分析Phishing页面敏感特征，提出一种防御Phishing攻击的Web页面检测算法。该算法通过分析Web页面的文档对象模型来提取Phishing敏感特征，使用BP神经网络检测页面异常程度，利用线性分类器判断该页面是否为Phishing页面。该算法成功过滤了Phishing页面，有效地阻止了Phishing攻击。

关键词: 网络钓鱼, Web页面检测算法, BP神经网络

CLC Number: 

• TP309