TY - GEN
T1 - Data-driven privacy analytics
T2 - 10th International Conference on Wireless Algorithms, Systems, and Applications, WASA 2015
AU - Wang, Rongrong
AU - Xue, Minhui
AU - Liu, Kelvin
AU - Qian, Haifeng
N1 - Publisher Copyright:
© Springer International Publishing Switzerland 2015.
PY - 2015
Y1 - 2015
N2 - Location-based Social Network (LBSN) services enable people to discover users nearby and establish the communication with them. WeChat as both LBSN and Online Social Network (OSN) application does not impose a real-name policy for usernames, leaving the users to choose how they want to be identified by nearby people. In this paper, we show the feasibility to stalk WeChat users in any city from any place in the world and in parallel examine the anonymity of those users. Based on previous studies, we develop an automated attacking methodology by using fake GPS location, smart phone emulation, task automation, and optical character recognition (OCR). We then study the prevalence and behavior of Anonymous and Identifiable WeChat users and correlate their anonymity with their behavior, especially for those who repeatedly query the People Nearby service, a feature that triggers WeChat to discover nearby people. By monitoring Wall Street for 7 days, we gather location information relevant to 3,215 distinct users and finally find that Anonymous users are largely less inhibited to be dynamic participants, as they query more and are more willing to move around in public. To the best of our knowledge, this is the first work that quantifies the relationship between user mobility and user anonymity. We expect our study to motivate better privacy design in WeChat.
AB - Location-based Social Network (LBSN) services enable people to discover users nearby and establish the communication with them. WeChat as both LBSN and Online Social Network (OSN) application does not impose a real-name policy for usernames, leaving the users to choose how they want to be identified by nearby people. In this paper, we show the feasibility to stalk WeChat users in any city from any place in the world and in parallel examine the anonymity of those users. Based on previous studies, we develop an automated attacking methodology by using fake GPS location, smart phone emulation, task automation, and optical character recognition (OCR). We then study the prevalence and behavior of Anonymous and Identifiable WeChat users and correlate their anonymity with their behavior, especially for those who repeatedly query the People Nearby service, a feature that triggers WeChat to discover nearby people. By monitoring Wall Street for 7 days, we gather location information relevant to 3,215 distinct users and finally find that Anonymous users are largely less inhibited to be dynamic participants, as they query more and are more willing to move around in public. To the best of our knowledge, this is the first work that quantifies the relationship between user mobility and user anonymity. We expect our study to motivate better privacy design in WeChat.
KW - Anonymity
KW - Location-based social networks
KW - Mobility
UR - https://www.scopus.com/pages/publications/84943620671
U2 - 10.1007/978-3-319-21837-3_55
DO - 10.1007/978-3-319-21837-3_55
M3 - 会议稿件
AN - SCOPUS:84943620671
SN - 9783319218366
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 561
EP - 570
BT - Wireless Algorithms, Systems, and Applications - 10th International Conference, WASA 2015, Proceedings
A2 - Xu, Kuai
A2 - Zhu, Haojin
PB - Springer Verlag
Y2 - 10 August 2015 through 12 August 2015
ER -