| At present,the public demand for Internet services is fast and convenient,but the Internet security has not been paid attention to.At the same time,because of the openness of the Internet,hackers have many opportunities.This will be a huge threat to people's privacy security,and even threaten people's normal life.In these attacks,the virus is loaded into JavaScript and malicious attacks are carried out through the vulnerability of the browser and its plug-ins.This attack has a serious impact on the security of the internet.Some measures have been put forward for this kind of attack.For example,the Trojan horse in Java Script is detected by feature code matching technology.But there is one sidedness in this way,that is,it is difficult to detect malicious code encrypted and confused.At the same time,because the method is based on static feature code,and the static feature library increases gradually,the detection performance will gradually decrease.If the dynamic monitoring method is used,the malicious operation can be judged,but if the amount of data is large,the detection efficiency will be greatly reduced.This paper proposes a Java Script script detection method,which can use classification to deal with malicious script of Java Script.The core idea of this method is to extract the vector rule of Java Script script content according to malicious behavior,and use support vector machine(SVM)for regression calculation to automatically determine whether the Java Script script is malicious code.This paper uses Visual Studio2015 tool to develop the system,and has the function of detecting Java Script code with encryption deformation confusion.The proposed detection system of high detection performance,and excellent detection of confusion encryption Java Script,and provides the browser,anti-virus engine interface,has higher application value,the thought of other types of judging whether the file is malicious code also provides greater ideas. |
PDF Full Text Request