Re-identification risk measurement estimation of a dataset

1. A method for estimating a re-identification risk to one of a plurality of subjects, each separately described in a respective subject profile in a dataset, the method comprising:

on a processor connected to a network and the dataset:

retrieving a population distribution determined for each of the one or more quasi-identifying fields in the subject profile, wherein the population distribution aps a value in the each quasi-identifying field to a probability of at least one of the plurality of subjects having this value;

assigning an information score to the each mapped value in the one or more quasi-identifying fields in the subject profile, wherein the information score is defined by a plurality of information bits in the mapped value;

aggregating the assigned information scores of all the mapped values in the one or more quasi-identifying fields in the subject profile;

estimating the re-identification risk to the one subject, based on

a plurality of re-identification bits, wherein the plurality of re-identification bits is a function [log 2 ] of a population,

an anonymity value, wherein the anonymity value is based on the aggregated assigned information scores and the plurality of re-identification bits, and

a re-identification metric for the subject based on the anonymity value; and

lowering the re-identification risk to the one subject by performing further de-identification of the dataset based upon the estimated re-identification risk.

2. The method of claim 1 , wherein the population distribution is a variable.

3. The method of claim 1 , further comprising a step of creating an aggregate result of a plurality of re-identification metrics for a plurality of subject profiles in the dataset.

4. A system for estimating re-identification risk to one of a plurality of subjects, each separately described in a respective subject profile in the a dataset, the system comprising:

a hardware processor connected to a network and the dataset, the processor is configured to: retrieve a population distribution determined for each of the one or more quasi-identifying fields described in the subject profile, wherein the population distribution maps a value in the each quasi-identifying field to a probability of at least one of the plurality of subjects having this value;

assign an information score to the each mapped value in the one or more quasi-identifying fields in the subject profile, wherein the information score is defined by a plurality of information bits in the mapped value; aggregate the assigned information scores of all the mapped values of-in the one or more quasi-identifying fields in the subject profile; and calculate estimate the re-identification risk to the one subject based on a plurality of re-identification bits, wherein the plurality of re-identification bits is a function [log 2 ] of a population, an anonymity value, wherein the anonymity value is based on the aggregated plurality of assigned information scores and the plurality of re-identification bits, and a re-identification metric for the subject based on the anonymity value; and lower the re-identification risk to the one subject by performing further de-identification of the dataset based upon the estimated re-identification risk.

5. The system of claim 4 , wherein the processor is further configured to creating an aggregate result of a plurality of re-identification metrics for a plurality of subject profiles in the dataset.

6. The system of claim 5 , wherein the aggregate result is a matrix.

7. The method of claim 3 , wherein the aggregate result is a matrix.

8. The method of claim 1 , wherein the aggregate of the assigned information scores is a summation of the plurality of assigned information bits of the information score of each value in the subject profile.

9. The method of claim 8 , wherein the summation is a total information value of all child records for the subject profile.

10. The method of claim 1 , wherein the subject profile comprises at least one record, and the method further comprising steps of:

aggregating the assigned information scores within each value in the at least one record;

aggregating the plurality of assigned information scores from values related to child tables associated with the record; and

aggregating the plurality of assigned information scores from values in the child tables.

11. The method of claim 1 , further comprising a step of selecting predefined values of the one or more quasi-identifying fields with the most information related to a given parent as determined by the plurality of assigned information bits.

12. The method of claim 1 , further comprising a step of calculating an arithmetic average information (u) of the plurality of assigned information bits in values of all elements related to the subject profile.

13. The method of claim 1 , wherein the re-identification metric is further defined by any of a value associated with anonymity, equivalence class size, and re-identification risk.

14. The method of claim 1 , further comprising steps of:

assigning a probability to each of the assigned information score defined by the plurality of information bits,

aggregating the assigned probabilities; and

evaluating the ability to unambiguously link a record in one dataset to identify a matching subject in another dataset.

15. The method of claim 1 , wherein the anonymity value is a metric measured in bits, where

if the anonymity value is greater than zero there are many subjects matching in the population,

if the anonymity is equal to zero the subjects is unique in the population, and

if the anonymity value is less than zero the subjects is unlikely to exist in the dataset or population.

16. The method of claim 15 , further comprising a step of generating a histogram from a plurality of calculated anonymity values to estimate a number of unique subjects in the dataset.

17. The system of claim 4 , wherein the population distribution is a variable.