Classification features for detecting server-side and client-side web attacks

Abstract

During last years, the number and cleverness of attacks against Web related applications are steadily growing as Web services become more popular. In this paper, we propose relevant classification features for detecting Web attacks targeting either server-side or client-side applications. Four kinds of features are provided: Request general features, Request content features, Response features and Request history features. Experimental studies carried on real1 and simulated htt p traffic including normal data and several attacks show the efficiency of our feature set in detecting Web related attacks. © 2008 Springer Science+Business Media, LLC.