一种欺骗网页检测判定算法  

Decision Algorithm for Detecting Deceptive Web Page

作　　者：余小燕[1] 陆全华[1] 

机构地区：[1]江苏畜牧兽医职业技术学院信息工程系,江苏泰州225300

出　　处：《成都大学学报（自然科学版）》2009年第4期332-335,共4页Journal of Chengdu University（Natural Science Edition）

摘　　要：欺骗网页与正常网页有着极其相似的特征,诱导并欺骗用户输入个人机密信息,严重威胁着电子商务正常运行.基于此,提出了一个欺骗网页检测算法.算法首先对网页进行分割,然后提取各个块的特征及其相对位置关系,得到网页特征关系图,并以此计算出两个网页之间的相似度.如果发现可疑网页与某个受保护网页之间的相似度小于预设值,判定该可疑网页为针对受保护网页的欺骗网页.实验结果表明,本算法可以有效检测欺骗网页,具有一定的实用性.Deceptive web pages have similarity with normal web pages and induce and cheat users to enter personal secret information, which poses a severe threat to the normal development of e-commerce. In light of this situation, an algorithm for detecting deceptive web pages was proposed, which was used to segment the web pages first. Then every block＇s characteristics and its relative position relationship were extracted to form characteristic relation diagram of web pages, by which amount of similarity between two pages was calculated. If the amount of similarity between a suspicious web page and the protected web page is less than the default value, then the suspicious web page can be judged a cheating page. The experimental findings indicate that this algorithm can effectively detect cheating web pages, which has a certain practicality.

关 键 词：欺骗网页检测 网页分割 相似度 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]