Detecting Vulnerabilities in Web Applications by Clustering Web Pages
Vanja Suhina, Stjepan Groš, Zoran Kalafatić · 2008
Abstract – In this paper, we propose a new approach to detecting vulnerabilities in Web applications. Majority of current Web application vulnerability scanners rely on detecting vulnerabilities by detecting common error messages or input vectors used in testing. The method we propose in this paper is based on detecting unusual behavior of a Web application. Differences between pages are analyzed by examining page structure, i.e. HTML elements. Variations from standard page structure could indicate raised errors in the Web application and could indicate a vulnerability. Issues that arise in building such a tool will be described here. I.