Paper ID
1601
Paper Type
full
Description
Consumer location tracking is becoming omnipresent on mobile devices, producing vast volumes of behavior-rich location trajectory data. However, this comes at a cost – potential invasion of consumer privacy. Existing approaches to privacy preservation are largely unsuited for these new data, not personalized, and difficult to interpret the trade-off between consumer privacy and data utility. We propose a personalized and interpretable framework to enable location data collectors to optimize the privacy-utility trade-off. Validating on a sample of nearly one million location trajectories from over 40,000 individuals, we find that high privacy risks indeed prevail in the absence of data obfuscation. Outperforming multiple baselines, the proposed framework significantly reduces the privacy risk with minimal decrease in an advertiser’s utility. As novel and powerful consumer location trajectory data become increasingly leveraged, we demonstrate its value and propose an interpretable framework to mitigate its risk while maximizing its value.
Recommended Citation
M Y, Meghanath; Li, Beibei; and Foutz, Ying Natasha Zhang, "Geo-Targeting, Privacy, and the Rise of Consumer Location Trajectories" (2019). ICIS 2019 Proceedings. 28.
https://aisel.aisnet.org/icis2019/cyber_security_privacy_ethics_IS/cyber_security_privacy/28
Geo-Targeting, Privacy, and the Rise of Consumer Location Trajectories
Consumer location tracking is becoming omnipresent on mobile devices, producing vast volumes of behavior-rich location trajectory data. However, this comes at a cost – potential invasion of consumer privacy. Existing approaches to privacy preservation are largely unsuited for these new data, not personalized, and difficult to interpret the trade-off between consumer privacy and data utility. We propose a personalized and interpretable framework to enable location data collectors to optimize the privacy-utility trade-off. Validating on a sample of nearly one million location trajectories from over 40,000 individuals, we find that high privacy risks indeed prevail in the absence of data obfuscation. Outperforming multiple baselines, the proposed framework significantly reduces the privacy risk with minimal decrease in an advertiser’s utility. As novel and powerful consumer location trajectory data become increasingly leveraged, we demonstrate its value and propose an interpretable framework to mitigate its risk while maximizing its value.