Differential Privacy Location Protection Scheme Based on Hilbert Curve
Algorithm 2
Anonymous data processing procedure.
Input: -privacy budget L-location data set D-non-location data set, -non-location data attribute collection, -continuous attribute data set, -discrete attribute data set, h-the height of the tree.
Output: anonymous data set T satisfying differential privacy protection.
(1)
Begin Procedure
(2)
;//privacy budget allocation
(3)
If the data belongs to location data L
(4)
For any element in set L
(5)
; //Q is the user’s query function, which has global sensitivity and adds Laplace noise to the location data
(6)
End For
(7)
Else If the data belongs to nonlocation data D
(8)
For any element P in set D, and the element satisfies //Attribute partition of nonlocation data
(9)
If is a continuous-valued attribute, then
(10)
; //Adding Laplace noise to nonlocation data with the continuous-valued attribute
(11)
Else If is a discrete-valued attribute, then
(12)
; //Laplace noise is added to nonlocation data with the discrete-valued attribute