Malicious Web Page Research And Realization Of The Automated Analysis And Recognition Technology

With the rapid development of Internet and WEB Service technology, WEBsecurity problem has become an important threat to public and network environmentsecurity. According to the latest CISSP statistics in2011, at present, WEB attacksaccounted for more than half of the all. Although Microsoft Corp. has developedNozzle and Zozzle system to protect IE from exploiting, the defense which be in thelight of web cross-site and other malicious page-attack, still have little effect. So theautomatic capture technology, identification, analysis and inverse system, at home andabroad, is an urgent need for solving the malicious web problems.In view of the present situation, especially which the web pages contain somemalicious code such as IE、FireFox and third party plug-ins vulnerabilities, the thesisdiscover a set of automated analysis solutions and develop a prototype program. Thisthesis,according to web pages data asynchronous transmission characteristics andbrowser vulnerabilities using code characteristic, put forward a new kind of technologysolutions: page data capture using browser plug-ins, dynamic JS script analysis, pageelement analytical correlation, stack overflow detection, heap allocation trackingtechnique to detect the malicious web page.Furthermore, the thesis has analyzed three types of vulnerability which makegreat influence to social in recent three years. According to the proposed scheme, wefinally realized WebGlance. WebGlance provide a convenient way of installing to theterminal user browser environment for real-time detection of malicious webpage. Inthe test process, the normal page and undocumented malicious pages were tested. Testresult show that webglance possess more detail representation about page depthanalysis and attack processes association, validate the scheme application value in webmalicious page analysis detection technology.