Methods, systems, articles of manufacture and apparatus to privatize consumer data

1. An apparatus comprising:

a consumer data acquirer to collect original data corresponding to (a) confidential information associated with consumers and (b) behavior information associated with the consumers; and

a data obfuscator to:

determine a degree to which the original data is to be obfuscated and a type of obfuscation to be applied to the original data based on (i) a degree of privacy and (ii) a degree of similarity between the original data and obfuscated data,

determine if a threshold level of the degree of privacy is met;

in response to the threshold level of the degree of privacy not being met, refrain from generating the obfuscation adjustments;

in response to the threshold level of the degree of privacy being met, generate obfuscation adjustments of the original data based on the determined degree to which the original data is to be obfuscated and the type; and

generate an obfuscation model based on the obfuscation adjustments.

2. The apparatus as defined in claim 1 , wherein the data obfuscator is to apply the obfuscation adjustments to the original data to generate obfuscated data and further including an obfuscation verifier to:

verify that the obfuscation adjustments exceed an obfuscation adjustment threshold value; and

verify that a similarity index value between first calculated data that is associated with the original data and second calculated data that is associated with the obfuscated data exceeds a similarity threshold.

3. The apparatus as defined in claim 1 , wherein the data obfuscator is to generate the obfuscation model by applying the obfuscation adjustments to (a) the confidential information and (b) the behavior information.

4. The apparatus as defined in claim 1 , further including:

a selector to select a set of the confidential information associated with at least one of the consumers; and

a calculator to calculate a likelihood of determining an identity of the at least one of the consumers based on the set of the confidential information and the behavior information, wherein the data obfuscator is to generate the obfuscation adjustments in response to the likelihood satisfying a threshold criteria.

5. The apparatus as defined in claim 1 , wherein the obfuscation model is an obfuscation data converter to be applied to another set of confidential information.

6. The apparatus as defined in claim 1 , wherein the obfuscation adjustments include noise to be applied to the original data.

7. The apparatus as defined in claim 6 , wherein the noise includes multiplicative Laplace noise or Poisson noise.

8. The apparatus as defined in claim 1 , further including a sorter to remove outliers from the original data.

9. The apparatus as defined in claim 1 , wherein the obfuscation adjustments include utilizing conditional entropy.

10. A method comprising:

collecting, by instructions executed with at least one processor, original data corresponding to (a) confidential information associated with consumers and (b) behavior information associated with the consumers;

determining, by instructions executed with the at least one processor, a degree to which the original data is to be obfuscated and a type of obfuscation to be applied to the original data based on (i) a degree of privacy and (ii) a degree of similarity between the original data and obfuscated data;

determining, by instructions executed with the at least one processor, if a threshold level of the degree of privacy is met;

in response to the threshold level of the degree of privacy not being met, refraining from generating the obfuscation adjustments;

in response to the threshold level of the degree of privacy being met, generating, by instructions executed with the at least one processor, obfuscation adjustments of the original data based on the degree to which the original data is to be obfuscated and the type; and

generating, by instructions executed with the at least one processor, an obfuscation model based on the obfuscation adjustments.

11. The method as defined in claim 10 , further including:

applying, by instructions executed with the at least one processor, the obfuscation adjustments to the original data to generate obfuscated data;

verifying, by instructions executed with the at least one processor, that the obfuscation adjustments exceed an obfuscation adjustment threshold value; and

verifying, by instructions executed with the at least one processor, that a similarity index value between first calculated data that is associated with the original data and second calculated data that is associated with the obfuscated data exceeds a similarity threshold.

12. The method as defined in claim 10 , wherein the generating of the obfuscation model includes applying the obfuscation adjustments to (a) the confidential information and (b) the behavior information.

13. The method as defined in claim 10 , further including:

selecting, by instructions executed with the at least one processor, a set of the confidential information associated with at least one of the consumers; and

calculating, by instructions executed with the at least one processor, a likelihood of determining an identity of the at least one of the consumers based on the set of the confidential information and the behavior information, wherein the obfuscation adjustments are generated in response to the likelihood satisfying a threshold criteria.

14. The method as defined in claim 10 , wherein the obfuscation model is an obfuscation data converter to be applied to another set of confidential information.

15. The method as defined in claim 10 , wherein the obfuscation adjustments include noise to be applied to the original data.

16. The method as defined in claim 15 , wherein the noise includes multiplicative Laplace noise or Poisson noise.

17. The method as defined in claim 10 , further including sorting, by instructions executed with the at least one processor, to remove outliers from the original data.

18. The method as defined in claim 10 , wherein the obfuscation adjustments include utilizing conditional entropy.

19. A non-transitory computer readable medium comprising instructions, which when executed, cause at least one processor to at least:

collect original data corresponding to (a) confidential information associated with consumers and (b) behavior information associated with the consumers;

determine a degree to which the original data is to be obfuscated and a type of obfuscation to be applied to the original data based on (i) a degree of privacy and (ii) a degree of similarity between the original data and obfuscated data;

determine if a threshold level of the degree of privacy is met;

in response to the threshold level of the degree of privacy not being met, refrain from generating the obfuscation adjustments;

in response to the threshold level of the degree of privacy being met, generate obfuscation adjustments of the original data based on the determined degree to which the original data is to be obfuscated and the type; and

generate an obfuscation model based on the obfuscation adjustments.

20. The non-transitory computer readable medium as defined in claim 19 , wherein the at least one processor is caused to:

apply the obfuscation adjustments to the original data to generate obfuscated data;

verify that the obfuscation adjustments exceed an obfuscation adjustment threshold value; and

verify that a similarity index value between first calculated data that is associated with the original data and second calculated data that is associated with the obfuscated data exceeds a similarity threshold.

21. The non-transitory computer readable medium as defined in claim 19 , wherein the at least one processor is caused to generate the obfuscation model by applying the obfuscation adjustments to (a) the confidential information and (b) the behavior information.

22. The non-transitory computer readable medium as defined in claim 19 , wherein the at least one processor is caused to:

select a set of the confidential information associated with at least one of the consumers; and

calculate a likelihood of determining an identity of the at least one of the consumers based on the set of the confidential information and the behavior information, wherein the obfuscation adjustments are generated in response to the likelihood satisfying a threshold criteria.

23. The non-transitory computer readable medium as defined in claim 19 , wherein the obfuscation model is an obfuscation data converter to be applied to another set of confidential information.

24. The non-transitory computer readable medium as defined in claim 19 , wherein the obfuscation adjustments include noise to be applied to the original data.

25. Thenon-transitory computer readable medium as defined in claim 24 , wherein the noise includes multiplicative Laplace noise or Poisson noise.

26. The non-transitory computer readable medium as defined in claim 19 , wherein the at least one processor is caused to sort the original data to remove outliers.

27. The non-transitory computer readable medium as defined in claim 19 , wherein the obfuscation adjustments include utilizing conditional entropy.

28. The apparatus as defined in claim 1 , wherein the data obfuscator is to determine the degree to which the original data is to be obfuscated based on a threshold accuracy of obfuscated data.

29. The apparatus as defined in claim 1 , wherein the data obfuscator is to determine the degree to which the original data is to be obfuscated based on obfuscated data being obfuscated to conceal at least one of the confidential information or the behavior information.

30. The apparatus as defined in claim 1 , wherein the data obfuscator is to determine if the threshold level is met by calculating a degree of privacy of obfuscated data and comparing the degree of privacy to the threshold level.

31. An apparatus comprising:

at least one memory;

instructions; and

a processor to execute the instructions to:

collect original data corresponding to (a) confidential information associated with consumers and (b) behavior information associated with the consumers,

determine a degree to which the original data is to be obfuscated and a type of obfuscation to be applied to the original data based on (i) a degree of privacy and (ii) a degree of similarity between the original data and obfuscated data,

determine if a threshold level of the degree of privacy is met;

in response to the threshold level of the degree of privacy not being met, refrain from generating the obfuscation adjustments;

in response to the threshold level of the degree of privacy being met, generate obfuscation adjustments of the original data based on the determined degree to which the original data is to be obfuscated and the type, and

generate an obfuscation model based on the obfuscation adjustments.

32. The apparatus as defined in claim 31 , wherein the processor is to:

apply the obfuscation adjustments to the original data to generate obfuscated data;

verify that the obfuscation adjustments exceed an obfuscation adjustment threshold value; and

verify that a similarity index value between first calculated data that is associated with the original data and second calculated data that is associated with the obfuscated data exceeds a similarity threshold.

33. The apparatus as defined in claim 31 , wherein the processor is to generate the obfuscation model by applying the obfuscation adjustments to (a) the confidential information and (b) the behavior information.

34. The apparatus as defined in claim 31 , wherein the processor is to:

select a set of the confidential information associated with at least one of the consumers; and

calculate a likelihood of determining an identity of the at least one of the consumers based on the set of the confidential information and the behavior information, wherein the obfuscation adjustments are generated in response to the likelihood satisfying a threshold criteria.

35. The apparatus as defined in claim 31 , wherein the obfuscation model is an obfuscation data converter to be applied to another set of confidential information.

36. The apparatus as defined in claim 31 , wherein the obfuscation adjustments include noise to be applied to the original data.

37. The-apparatus as defined in claim 36 , wherein the noise includes multiplicative Laplace noise or Poisson noise.

38. The apparatus as defined in claim 31 , wherein the processor is to sort the original data to remove outliers.

39. The apparatus as defined in claim 31 , wherein the obfuscation adjustments include utilizing conditional entropy.