PLDP: Personalized Local Differential Privacy for Multidimensional Data Aggregation

Algorithm 1

Personalized multiple optimized unary encoding (PMOUE).

(i)	Input: X = {, , , }: the raw data record with m unemptied attributes
	: the privacy budget that allocated to attribute
	: the number of the candidate values of each attribute
	Output: Y = {, , , }: the perturbed data record with m attributes
(1)	Encoding:
(2)	Perturbation:
(3)	return Y = .