Fight against Misbehaving Users on Online Social Networks
People
Hanqiang Cheng, School of Computer Science, McGill University
Xinyu Xing, School of Computer Science, Georgia Institute of Technology
Xue Liu, Associate Professor, School of Computer Science, McGill University
Research Statement
Online social networks (OSNs), such as Facebook and Twitter, are becoming increasingly influential communication platforms connecting billions of people all around the world. As of August 2014, Facebook had over 1.2 billion monthly active users who spent about 40 minutes on Facebook per day on average. Twitter had over 200 million active users who sent 400 million tweets every day (i.e., messages posted on Twitter). An ecosystem on top of OSNs is rapidly fostered and everyone involved benefits from the ecosystem. Normal people mainly use OSNs to hear the latest news and interact with friends. Celebrities and public figures embrace OSNs to spread news and share ideas. Companies and advertisers leverage OSNs to promote brands and products. OSNs providers utilize the large user base of OSNs and valuable social data for various business purposes such as online advertisements.
Due to the open nature for information generation and the effective mechanism of information diffusion, an increasing number of misbehaving users abuse OSNs by disseminating inappropriate content, including malicious spam content as well as sexually explicit content (a.k.a. adult content). The prevalence of misbehaving users is not only harmful to the prosperous ecosystem of OSNs, but also exposes users to various attacks such as phishing, malware and fraud. Particularly, it is extremely dangerous for children who are vulnerable to the inappropriate content distributed by misbehaving users. This thesis mainly focuses on two types of misbehaving users who are the mostly complained about by the general public: (1) misbehaving users who disseminate malicious spam content, and (2) misbehaving users who disseminate sexually explicit content.
Publications
Hanqiang Cheng, Xinyu Xing, Xue Liu and Qin Lv. "ISC: An Iterative Social based Classifier for Adult Account Detection on Twitter", TKDE 2014.
Xinyu Xing, Yu-Li Liang, Sui Huang, Hanqiang Cheng, Richard Han, Qin Lv, Xue Liu, Shivakant Mishra and Yi Zhu, "Scalable misbehavior detection in online video chat services", KDD 2012
Xinyu Xing, Yu-Li Liang, Hanqiang Cheng, Jianxun Dang, Rechard Han, Xue Liu, Qin Lv and Shivakant Mishra, "SafeVchat: Detecting Obscene Content and Misbehaving Users in Online Video Chat Services", WWW 2011.
Xinyu Xing, Jianxun Dang, Richard Han, Xue Liu and Shivakant Mishra, "Intrusions into Privacy in Video Chat Environments: Attacks and Countermeasures",
Technical Report CU-CS-1068-10, 2010.