DETECTION OF APP RANKING FRAUD AND MALWARE IN GOOGLE PLAY STORE USING MACHINE LEARNING TECHNIQUES

Abstract

The rise of mobile applications has significantly contributed to the growth of digital ecosystems, with platforms like Google Play hosting millions of apps used by billions of users worldwide. However, the openness of such marketplaces has made them a prime target for malicious activities such as search rank fraud and malware distribution. Search rank fraud involves deceptive techniques including fake reviews, inflated ratings, and manipulated install counts, which artificially boost an app’s ranking and mislead genuine users. Malware, on the other hand, poses severe risks by stealing sensitive information, tracking user activity, or compromising device functionality. This paper introduces Fair Play, a detection framework designed to identify both search rank fraud and malware in Google Play. The system integrates behavioural analysis, co-review graph modelling, temporal tracking, and linguistic evaluation to uncover fraudulent patterns. By combining static and dynamic analysis techniques with supervised machine learning, Fair Play achieves high accuracy in classifying malicious and fraudulent applications. Experimental evaluation demonstrates that the system can detect fraudulent apps with more than 97% accuracy and malware with over 95% accuracy, thereby enhancing marketplace trust and ensuring user safety.