Support Personalized Weighted Local Differential Privacy Skyline Query

Algorithm 1

Personalized privacy budget allocation.

	Input: : each data record has attributes; : denotes the overall privacy budget.
	Output: : the privacy budget allocated to each attribute .
(1)	Construct the judgment matrix of the data set according to equation (4)
(2)
(3)	According to equation (5), we obtain the consistency index
(4)	According to equation (6), we calculate CR
(5)	if then
(6)	Denotes a failure of the consistency checking
(7)	else
(8)	By passing the consistency check, the feature vector is calculated based on
(9)	end if
(10)	: normalization
(11)	Attributes of the standardized data record
(12)
(13)
(14)
(15)	According to equation (10), we can get
(16)
(17)	return ;