期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：3
页码：5723
DOI：10.15680/IJIRCCE.2017.0503149
出版社：S&S Publications
摘要：Facebook applications are very attractive now a days with 20 million installs a calendar day, third-partyapps are a main reason for the reputation and addictiveness of Facebook. Unfortunately, hackers have take in thepotential of using apps for spreading malware and spam. The problem is before now considerable, as we find that atleast 13% of apps in datasets are malicious. So the research community has focused on detecting malicious posts andcampaigns. In a Facebook application, can we determine if it is malicious? Our explanation involvement is indeveloping FRAppE Facebook Rigorous Application Evaluator possibly the first tool focused on detecting maliciousapps on Facebook. To extend FRAppE, we use information grouped by scrutinizing the posting behaviour of 112KFacebook apps seen transversely 2.3 million users on Facebook. First, we recognize a set of types that help usdistinguish malicious apps from benign ones. We find that malicious apps often distribute names with other apps, andthey normally request less permission than benign apps. Second, leveraging these distinguishing features, we show thatFRAppE can identify malicious apps with 99.6% accuracy, with no false positives and a high true positive rate(95.8%). Finally, we walk around the ecosystem of malicious Facebook apps and recognize mechanisms that these appsuse to propagate.
关键词：Facebook apps; malicious; online social networks; spam.
