作者: Engin Kirda , Richard A. Kemmerer , Christopher Kruegel , Greg Banks , Giovanni Vigna
DOI:
关键词:
摘要: Spyware is rapidly becoming a major security issue. programs are surreptitiously installed on user's workstation to monitor his/her actions and gather private information about behavior. Current antispyware tools operate in way similar traditional antivirus tools, where signatures associated with known spyware checked against newly-installed applications. Unfortunately, these techniques very easy evade by using simple obfuscation transformations. This paper presents novel technique for detection that based the characterization of spywarelike The tailored popular class applications use Internet Explorer's Browser Helper Object (BHO) toolbar interfaces browsing Our uses composition static dynamic analysis determine whether behavior BHOs toolbars response simulated browser events should be considered malicious. evaluation our representative set samples show it possible reliably identify malicious components an abstract behavioral characterization.