Locational privacy-preserving distance computations with intersecting sets of randomly labeled grid points

Area of Science:

• Computer Science
• Data Privacy
• Geospatial Analysis

Background:

• A novel privacy-preserving distance computation method, Intersecting Sets of Grid Points (ISGP), is introduced.
• ISGP utilizes randomly labeled grid points and hash values to mask spatial data coordinates.
• This method addresses challenges in computing distances when temporal location data is not simultaneously available.

Purpose of the Study:

• To introduce and study the ISGP method for privacy-preserving distance computations.
• To demonstrate the applicability of ISGP for handling geo-referenced data and record linkage.
• To evaluate the accuracy and efficiency of ISGP through simulations and real-world data.

Main Methods:

• ISGP is based on intersecting sets of randomly labeled grid points.
• Spatial coordinates are replaced by sets of hash values for privacy.
• The method was implemented in R and tested using large-scale real-world hospital and residential location data.

Main Results:

• ISGP achieves highly accurate approximations of distances between masked spatial data.
• Simulation studies demonstrated up to 99% accuracy with optimized parameters.
• The method proved effective in handling geo-referenced datasets and facilitating privacy-preserving record linkage.

Conclusions:

• ISGP is a highly accurate and efficient method for privacy-preserving distance computations in microdata.
• The method has low computational and storage requirements.
• ISGP offers a valuable tool for secure handling of geo-referenced data and has been applied in real-world scenarios.