Android Malware Detection Using Fine-Grained Features

摘要: Nowadays, Android applications declare as many permissions possible to provide more function for the users, which also poses severe security threat them. Although malware detection methods based on have been developed, they are ineffective when malicious few dangerous or declared by similar with those benign applications. This limitation is attributed use of too information classification. We propose a new method named fine-grained permission (FDP) detecting applications, gathers features that better represent difference between and Among these features, feature applied in components proposed first time. evaluate 1700 1600 demonstrate FDP achieves TP rate 94.5%. Furthermore, compared other related approaches, can detect families only requires 15.205 s analyze one application average, demonstrates its applicability practical implementation.