RETRACTED ARTICLE: A novel permission ranking system for android malware detection—the permission grader

Abstract

Android is a profound, vanguard mobile operating system of the contemporary era. The quantity of mobile phone emptor dependent on Android platform is rising expeditiously, which expands its prominence everywhere throughout the world. The facts demonstrate that there are different sides to everything, with the remarkable achievement of the Android; attacks on Android operating system have been on the rise, as there are a lot of apps on the Internet that encompass malware. Malware is a segment of code composed with the aim of hurting a gadget or stealing the information in it. The proposed Permission Grading System performs static analysis of the apps to extract the requested permissions and to identify the permissions that are unique to malware and benign apps, by calculating the contribution of each of the permission. This helps identify the risk of the permissions that are requested by the apps. The results show an increase of about 20% in the detection of malware, with True Positive Rate values more than 0.85 and False Positive Rate values nearly fall below 0.03. These values are improved on using the familiar Term Frequency—Inverse Document Frequency weighting after the identification of unique permissions. This has led to achieve a True Positive Rate of more than 0.90 and False Positive Rate values were only 0.01.