A Survey on Differentially Private Techniques for Spatial Databases

PhD Qualifying Examination


Title: "A Survey on Differentially Private Techniques for Spatial Databases"

by

Mr. Maocheng LI


Abstract:

Spatial data (i.e., locations, trajectories, and etc.) is ubiquitous in our 
daily life. When we order a taxi from a car-hailing platform (e.g., Uber or 
Lyft), the origin and destination of our journey, together with the entire 
trace of GPS locations during the trip are sent to the platform. While we 
recognize the usefulness of sharing spatial data, the privacy issue has become 
a major concern. Recent research shows that more than 80% of individuals could 
be uniquely identified from their released trajectories, even if we remove 
Personally Identifiable Information (PII, e.g., the national ID) associated 
with the location data. In this survey, we review important differential 
privacy (DP) related techniques in spatial databases. DP is considered as the 
de-facto golden standard for privacy protection in database management. This 
survey comprises three parts: i) we review necessary background related to DP, 
and its more recent variants: local-DP (LDP) and metric LDP; ii) we survey 
representative research works on location data and trajectories data, 
respectively; iii) we review DP techniques for Location-based services (LBS) 
applications data, with a focus on a differentially-private framework, kSwitch. 
We conclude the survey by discussing open questions and directions for future 
works.


Date:			Monday, 9 January 2023

Time:                  	4:00pm - 6:00pm

Zoom Meeting:		https://hkust.zoom.us/j/9419119233

Committee Members:	Prof. Lei Chen (Supervisor)
  			Prof. Raymond Wong (Chairperson)
 			Prof. Xiaofang Zhou
 			Dr. Hao Liu (EMIA)


**** ALL are Welcome ****