A Novel Approach to Detect Android Malware

Android is the world's most popular and the largest installed base of any mobile platform. It has gained huge popularity among Smartphones and is growing very fast because it gives its users a world class platform for creating apps and games and allows them to be distributed instantly. Secondly, it also offers ample free third party applications to be downloaded and installed from Google Play, the premier marketplace for selling and distributing Android apps. Android openness has made it a favourite for users and developers alike. Many apps are getting downloaded by the user from play store every month. However due to this, the number of harmful apps in the form of malwares getting downloaded are also increasing. These malwares performs the various activities behind the scene, such as stealing various sensitive information of the users and signing up victims to subscription services. As a result of this, users are getting affected and their privacy gets compromised. As developers are also free to develop and publish their own creation in a play store without undergoing any scrutiny of their apps, they tend to take the advantage of user's inability to analyze the risk of such apps.This paper proposes a system which would help the users in analysing and removing such harmful apps and thereby protecting their security and privacy. This is achieved by analysing the various permissions used by an application that it has requested during installation. The overall process of analysing apps is done using clustering and classification techniques. The major objective of the proposed system is to detect and remove the malwares that are present in the user's Android device.