Specifically? the new method implemente Principle of the by INSEE is that of “random keys” ( cell key method ). It is based on the random assignment of a “key” (in concrete terms? a number between 0 and 1) to each individual (person? household? company? etc.). This assignment is carried out once and for all: each individual then always keeps the same key for all subsequent processing. When disseminating a data table? the individual keys are combined with a “transition matrix” to determine? based on the original (undisturbed) Principle of the value? the deviation to be applied to obtain the perturbed value that will be disseminated. This method is presented in more detail in a methodological sheet .
Explanation by example figure
Data on job seekers are disseminate QPV by QPV new zealand phone number library and broken down by gender and age group (under 26? 26 to under 50? 50 and over). The confidentiality threshold governing their dissemination is 5: by applying the confidentiality management method? the aim is therefore to limit the possibility for a data user to know with certainty? from the information disseminated on job seekers by gender and age group? whether information concerns fewer than 5 individuals.
In the example in Figure 2 ? which the more sites you have corresponds to a fictitious QPV? there are only 3 male job seekers under 26 years old. The direct dissemination of this information would therefore contravene the source’s dissemination rules.
Figure – Example of application of the random
In this context? the random key method is applied as follows:
1. In the initial (unperturbed) data aero leads file of job seekers? there are 3 men under 26 in the fictitious QPV (subfigure a) .
2. Each individual in the file has a perfectly random “individual key” (a number between 0 and 1). This is particularly the case for the 3 men under 26 in the box (subfigure b) .
