Statistical Data Learning Under Privacy Constraints

• US 20140122385A1
• Filed: 11/01/2012
• Published: 05/01/2014
• Est. Priority Date: 11/01/2012
• Status: Active Grant

• Current Assignee: GM Global Technology Operations LLC (General Motors Company)
• Original Assignee: General Motors LLC (General Motors Company)
• Inventors: Hecht, Ron M., Hillel, Aharon Bar, Lavi, Nadav
• Granted Patent: US 9,037,520 B2
• Time in Patent Office: Days
• US Class Current: 706/12
• CPC Class Codes: G06F 21/6254 by anonymising data, e.g. decorrelating personal data from the owner's identification
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G10L 15/197 Probabilistic grammars, e.g. word n-grams

Abstract

A computer-implemented method is provided for statistical data learning under privacy constraints. The method includes: receiving, by a processor, a plurality of pieces of statistical information relating to a statistical object and aggregating, by the processor, the plurality of pieces of statistical information so as to provide an estimation of the statistical object. Each piece of statistical information includes an uncertainty variable, the uncertainty variable being a value determined from a function having a predetermined mean. The number of pieces of statistical information aggregated is proportional to the reliability of the estimation of the statistical object.

Claims

1. A computer-implemented method for statistical data learning under privacy constraints, the method comprising:

receiving, by a processor, a plurality of pieces of statistical information relating to a statistical object, wherein each piece of statistical information includes an uncertainty variable, the uncertainty variable being a value determined from a function having a predetermined mean; and

aggregating, by the processor, the plurality of pieces of statistical information so as to provide an estimation of the statistical object, wherein the number of pieces of statistical information aggregated is proportional to the reliability of the estimation of the statistical object.

2. The method of claim 1, wherein the function is a Gaussian noise distribution.

3. The method of claim 1, wherein the predetermined mean is zero.

4. The method of claim 1, wherein the statistical object is a frequency or probability of use of a word in a language model, and each piece of statistical information is a frequency or probability of use of the word in the language model from a single entity.

5. The method of claim 1, further comprising:

before the receiving step, querying, by the processor, a plurality of entities for the plurality of pieces of statistical information relating to the statistical object.

6. The method of claim 1, wherein the computer-implemented method is executed according to computer-executable instructions at a centralized information aggregator, and the plurality of pieces of statistical information are received from a plurality of mobile communication units.

7. The method of claim 6, wherein the mobile communication units are cellular telephones.

8. The method of claim 6, wherein the mobile communications units are vehicle telematics units, and the centralized information aggregator is a call center.

9. The method of claim 8, further comprising:

transmitting, by the call center, results of the aggregating step to a plurality of vehicle telematics units.

10. The method of claim 1, wherein the computer-implemented method is executed according to computer-executable instructions at a vehicle telematics unit, and the plurality of pieces of statistical information are received from other vehicle telematics units.

11. The method of claim 1, wherein the statistical object is one of an average vehicle speed, mileage, or revolutions per minute (RPM).

12. A non-transitory computer-readable medium having computer-executable instructions stored thereon for statistical data learning under privacy constraints, the computer-executable instructions, when executed by a processor, causing the following steps to be performed:

receiving a plurality of pieces of statistical information relating to a statistical object. wherein each piece of statistical information includes an uncertainty variable, the uncertainty variable being a value determined from a function having a predetermined mean; and

aggregating the plurality of pieces of statistical information so as to provide an estimation of the statistical object, wherein the number of pieces of statistical information aggregated is proportional to the reliability of the estimation of the statistical object.

13. The computer-readable medium of claim 12, wherein the function is a Gaussian noise distribution.

14. The computer-readable medium of claim 12, wherein the predetermined mean is zero.

15. The computer-readable medium of claim 12, wherein the statistical object is a frequency or probability of use of a word in a language model, and each piece of statistical information is a frequency or probability of use of the word in the language model from a single entity.

16. The computer-readable medium of claim 12, wherein the computer-executable instructions, when executed by a processor, further cause the following step to be performed:

before the receiving step, querying a plurality of entities for the plurality of pieces of statistical information relating to the statistical object.

17. The computer-readable medium of claim 12, wherein the plurality of pieces of statistical information are received from a plurality of mobile communication units.

18. The computer-readable medium of claim 17, wherein the mobile communications units are vehicle telematics units, and wherein the computer-readable medium is part of a centralized information aggregator.

19. The computer-readable medium of claim 18, wherein the computer-executable instructions, when executed by a processor, further cause the following step to be performed:

transmitting results of the aggregating step to a plurality of vehicle telematics units.

20. The computer-readable medium of claim 17, wherein the computer-readable medium is part of a vehicle telematics unit, and the plurality of pieces of statistical information are received from other vehicle telematics units.