Malicious Applications Detection by Analyzing Manifest Files

Abstract

With the rapid rise in the use of Android worldwide, the harm done due to malicious applications is increasing exponentially. Due to this the data of the Android users is not secure, and we need a system to detect the maliciousness of the applications being downloaded. The project being developed here is a solution to this problem as it detects whether an application is malicious or benign, this will not only provide data security but will also enhance the performance of the user's device. There are a number of systems like DroidMat and signature-based systems, already available in the market that detect malicious applications. But these systems detect that on the base of signatures and by parsing smali files which not only increases the processing time but also does not provide accurate results. The proposed system here that is Malicious applications detection using permissions retrieval detects whether an application is malicious or not based on the keywords present in the manifest file. This method does not affect the performance of the device in any way and gives us better throughput.