Fine-grained Detection And Research Based On Multi-feature For Android Pirate Applications

With the rapid development of the Android system,more and more developers have developed hundreds of millions Android applications on the Android application downloading channels.Because of the poor supervision of application downloading channels,a large number of Android pirates are appeared in downloading channels.Most pirated Android applications have privacy theft,stealing traffic consumption,advertising popups,malicious hijacking and a series of hooliganism.On the one hand,it destroys the good ecological environment of the application market,on the other hand it poses a serious security threat to the developers and users.The detection of Android’s pirated application is of great significance to standardizing the application of the ecological environment for Android and protecting the rights and interests of the developers and users.In the aspect of Android piracy application similarity detection,there are problems in code based detection methods,such as decompiling and fast batch detection.And the detection method based on resource files has high false positive rate.In this thesis,a new method of similarity detection for pirated applications based on the features of Android multidimensional metadata is proposed.It can effectively and quickly detect Android piracy application from the massive Android applications.In the aspect of Android piracy detection,most of the existing researches focused on the detection of piracy application,and there is a lack of research on the malicious behavior of piracy.In this thesis,a set of dynamic and static combination of Android pirated application malicious behavior detection scheme is proposed.This method can accurately identify malicious applications in Android pirated applications with malicious behavior.In the application of the malicious classification of Android piracy,the existing researches on the malware classification exists misjudgments and other problems.This thesis proposes and designs a feature combination classification algorithm for malicious applications,which can effectively classify the malicious behaviors of Android piracy,and is more conducive to standardize the download channel of Android application market.