Abstract
Fake practices in Google Play, the most well known Android application market, fuel search rank maltreatment and malware multiplication. To recognize malware, past work has zeroed in on application executable and consent examination. In this paper, we present FairPlay, an original framework that finds and use follows left behind by fraudsters, to recognize both malware and applications exposed to look through rank misrepresentation. FairPlay corresponds to audit exercises and exceptionally joins identified survey relations with phonetic and social signs gathered from Google Play application information (87K applications, 2.9M audits, and 2.4M commentators, gathered over a large portion of a year), to distinguish dubious applications. FairPlay accomplishes more than 95% exactness in grouping best quality level datasets of malware, fake and authentic applications. We show that 75% of the distinguished malware applications participate in search rank extortion. FairPlay finds many deceitful applications that at present dodge Google Bouncer's recognition innovation. FairPlay likewise helped the disclosure of over 1,000 audits, detailed for 193 applications, that uncover another kind of "coercive" survey crusade: clients are pestered into composing positive surveys and introduce and survey other applications.