Identification and Classification of Malicious Third-Party Applications on Online Social Networks

Author

M. Sreedhar Reddy, Haricharan, Mahesh Kumar, B Vaishnavi

Abstract

 Online Social Networks (OSNs) support a large ecosystem of third-party applications that improve user engagement but can also introduce security and privacy risks. Malicious applications may abuse granted permissions to access private information, perform clickjacking, distribute spam, and propagate additional malicious applications. This paper presents FRAppE (Facebook's Rigorous Application Evaluator), together with its lightweight FRAppE Lite variant, for identifying malicious third-party applications. The approach profiles applications using on-demand features and aggregate cross-user behavioral features and applies Naïve Bayes and Support Vector Machine (SVM) classification. The source study reports evaluation on 111,000 applications associated with more than 91 million wall posts, with FRAppE Lite reporting 99.0% accuracy and Full FRAppE reporting 99.5% accuracy.

Keywords

Online Social Networks; Malicious Application Detection; FRAppE; Naïve Bayes; Support Vector Machine; OAuth 2.0; Application Profiling; App-Nets; Social Network Security.

Full Text:

Download Paper PDF

References

1. Facebook Open Graph API Documentation, 2016.
2. MyPageKeeper Security Platform, IEEE TPDS, 2012.
3. Chia et al., “Is this app safe? A large scale study on application permissions and risk signals,” Proc. WWW, 2012.
4. Gao et al., “Detecting and characterizing social spam campaigns,” Proc. ACM IMC, 2010.
5. S. Rahman et al., “Efficient and Scalable Socware Detection in OSNs,” USENIX Security, 2012.