TY - GEN
T1 - Privacy-preserving inference of social relationships from location data
T2 - 23rd ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2015
AU - Shahabi, Cyrus
AU - Fan, Liyue
AU - Nocera, Luciano
AU - Xiong, Li
AU - Li, Ming
N1 - Publisher Copyright:
© 2015 ACM.
PY - 2015/11/3
Y1 - 2015/11/3
N2 - Social relationships between people, e.g., whether they are friends with each other, can be inferred by observing their behaviors in the real world. Thanks to the popularity of GPS-enabled mobile devices or online services, a large amount of high-resolution location data becomes available for such inference studies. However, due to the sensitivity of loca- tion data and user privacy concerns, those studies cannot be largely carried out on individually contributed data without privacy guarantees. Furthermore, we observe that the actual location may not be needed for social relationship studies, but rather the fact that two people met and some statistical properties about their meeting locations, which can be com- puted in a private manner. In this paper, we envision an extensible framework, dubbed Privacy-preserving Location Analytics and Computation Environment (PLACE), which enables social relationship studies by analyzing individually generated location data. PLACE utilizes an untrusted server and computes several building blocks to support various so- cial relationship studies, without disclosing location infor- mation to the server and other untrusted parties. We present PLACE with three example social relationship studies which utilize four privacy-preserving blocks with encryption and differential privacy primitives. The successful realization of PLACE will facilitate private location data acquisition from individual devices, thanks to the strong privacy guarantees, and will enable a wide range of applications.
AB - Social relationships between people, e.g., whether they are friends with each other, can be inferred by observing their behaviors in the real world. Thanks to the popularity of GPS-enabled mobile devices or online services, a large amount of high-resolution location data becomes available for such inference studies. However, due to the sensitivity of loca- tion data and user privacy concerns, those studies cannot be largely carried out on individually contributed data without privacy guarantees. Furthermore, we observe that the actual location may not be needed for social relationship studies, but rather the fact that two people met and some statistical properties about their meeting locations, which can be com- puted in a private manner. In this paper, we envision an extensible framework, dubbed Privacy-preserving Location Analytics and Computation Environment (PLACE), which enables social relationship studies by analyzing individually generated location data. PLACE utilizes an untrusted server and computes several building blocks to support various so- cial relationship studies, without disclosing location infor- mation to the server and other untrusted parties. We present PLACE with three example social relationship studies which utilize four privacy-preserving blocks with encryption and differential privacy primitives. The successful realization of PLACE will facilitate private location data acquisition from individual devices, thanks to the strong privacy guarantees, and will enable a wide range of applications.
KW - Location privacy
KW - Social relationship
UR - http://www.scopus.com/inward/record.url?scp=84961203044&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84961203044&partnerID=8YFLogxK
U2 - 10.1145/2820783.2820880
DO - 10.1145/2820783.2820880
M3 - Conference contribution
AN - SCOPUS:84961203044
T3 - GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems
BT - 23rd ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2015
A2 - Huang, Yan
A2 - Ali, Mohamed
A2 - Sankaranarayanan, Jagan
A2 - Renz, Matthias
A2 - Gertz, Michael
PB - Association for Computing Machinery
Y2 - 3 November 2015 through 6 November 2015
ER -