GENERATING ANONYMOUS DATA FROM WEB DATA

• US 20150278157A1
• Filed: 03/25/2014
• Published: 10/01/2015
• Est. Priority Date: 03/25/2014
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Inventors: SRIVASTAVA, Ashok N.
• Granted Patent: US 9,866,454 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation
• CPC Class Codes: H04L 67/535 Tracking the activity of the user network monitoring arrangements H04L43/00; recording of computer activity G06F11/34

Abstract

A device receives web data, associated with user devices, that is generated based on interactions of the user devices with a network and one or more content provider devices. The device removes erroneous or objectionable web data from the web data to generate a subset of the web data, and categorizes the subset of the web data by assigning categories to the subset of the web data. The device performs an empirical estimation of the categorized subset of the web data to generate empirical estimations. The device performs a simulation of the empirical estimations to generate synthetic data that corresponds to the web data and removes private information relating to the user devices and users of the user devices, and stores the synthetic data in a storage device.

Claims

1. A method, comprising:

receiving, by a device, web data associated with user devices,

the web data being generated based on interactions of the user devices with a network and one or more content provider devices;

removing, by the device, erroneous or objectionable web data from the web data to generate a subset of the web data;

categorizing, by the device, the subset of the web data by assigning categories to the subset of the web data;

performing, by the device, an empirical estimation of the categorized subset of the web data to generate empirical estimations;

performing, by the device, a simulation of the empirical estimations to generate synthetic data that corresponds to the web data and removes private information relating to the user devices and users of the user devices; and

storing, by the device, the synthetic data in a storage device.

2. The method of claim 1, further comprising:

presenting, for display, the synthetic data to a device associated with at least one of the one or more content provider devices.

3. The method of claim 1, further comprising:

determining whether statistical properties and privacy, associated with the web data, are preserved in the synthetic data;

outputting the synthetic data when the statistical properties and the privacy, associated with the web data, are preserved in the synthetic data; and

re-performing the simulation of the empirical estimations to generate new synthetic data when the statistical properties or the privacy, associated with the web data, is not preserved in the synthetic data.

4. The method of claim 1, where the synthetic data includes similar statistical properties as the web data, without the private information from the web data.

5. The method of claim 1, where the simulation of the empirical estimations includes a Monte Carlo simulation of the empirical estimations.

6. The method of claim 1, where the web data includes one or more of:

clickstream data that includes information associated with portions of content, provided by the one or more content provider devices, that are selected via the user devices,

location data that includes information associated with locations of the user devices when the content is accessed by the user devices,

time data that includes information associated with times when the user devices access the content, or

network data that includes information associated with network resources utilized by the user devices to access the content.

7. The method of claim 1, where the empirical estimation of the categorized subset of the web data includes an empirical estimation of joint distributions of the categorized subset of the web data.

8. A device, comprising:

one or more processors to;

receive web data associated with user devices,

the web data being generated based on interactions of the user devices with a network and a plurality of content provider devices;

remove erroneous or objectionable web data from the web data to generate a subset of the web data;

categorize the subset of the web data by assigning categories to the subset of the web data;

perform an empirical estimation of the categorized subset of the web data to generate empirical estimations;

perform a simulation of the empirical estimations to generate synthetic data that corresponds to the web data and removes private information relating to the user devices and users of the user devices;

store the synthetic data in a storage device; and

output the synthetic data.

9. The device of claim 8, where, when outputting the synthetic data, the one or more processors are further to:

present, for display, the synthetic data to a user of the device or to a particular device associated with the plurality of content provider devices.

10. The device of claim 8, where the one or more processors are further to:

determine whether statistical properties and privacy, associated with the web data, are preserved in the synthetic data; and

re-perform the simulation of the empirical estimations to generate new synthetic data when the statistical properties or the privacy, associated with the web data, is not preserved in the synthetic data, and

where, when outputting the synthetic data, the one or more processors are further to;

output the synthetic data when the statistical properties and the privacy, associated with the web data, are preserved in the synthetic data; and

11. The device of claim 8, where the synthetic data includes similar statistical properties as the web data without the private information from the web data.

12. The device of claim 8, where the simulation o the empirical estimations includes a Monte Carlo simulation of the empirical estimations.

13. The device of claim 8, where the web data includes one or more of:

clickstream data that includes information associated with portions of content, provided by the plurality of content provider devices, that are selected via the user devices,

location data that includes information associated with locations of the user devices when the content is accessed by the user devices,

time data that includes information associated with times when the user devices access the content, or

network data that includes information associated with network resources utilized by the user devices to access the content.

14. The device of claim 8, where the empirical estimation of the categorized subset of the web data includes an empirical estimation of joint distributions of the categorized subset of the web data.

15. A computer-readable medium for storing instructions, the instructions comprising:

one or more instructions that, when executed by one or more processors of a device, cause the one or more processors to;

receive web data associated with user devices,

the web data being generated based on interactions of the user devices with a network and one or more content provider devices,

the web data including private information regarding the user devices and/or users of the user devices;

remove erroneous or objectionable web data from the web data to generate a subset of the web data;

categorize the subset of the web data by assigning categories to the subset of the web data;

perform an empirical estimation of the categorized subset of the web data to generate empirical estimations;

perform a simulation of the empirical estimations to generate synthetic data that statistically corresponds to the web data and removes the private information from the web data; and

store the synthetic data in a storage device.

16. The computer-readable medium of claim 15, where the instructions further comprise:

one or more instructions that, when executed by the one or more processors, cause the one or more processors to;

present, for display, the synthetic data to a user of the device or to a particular device associated with the one or more content provider devices.

17. The computer-readable medium of claim 15, where the instructions further comprise:

one or more instructions that, when executed by the one or more processors, cause the one or more processors to;

determine whether statistical properties and privacy, associated with the web data, are preserved in the synthetic data;

output the synthetic data in the data structure when the statistical properties and the privacy, associated with the web data, are preserved in the synthetic data; and

re-perform the simulation of the empirical estimations to generate new synthetic data when the statistical properties or the privacy, associated with the web data, is not preserved in the synthetic data.

18. The computer-readable medium of claim 15, where the synthetic data includes similar statistical properties as the web data without the private information from the web data.

19. The computer-readable medium of claim 15, where the simulation of the empirical estimations includes a Monte Carlo simulation of the empirical estimations.

20. The computer-readable medium of claim 15, where the web data includes one or more of:

clickstream data that includes information associated with portions of content, provided by the one or more content provider devices, that are selected via the user devices,

location data that includes information associated with locations of the user devices when the content is accessed by the user devices,

time data that includes information associated with times when the user devices access the content, or

network data that includes information associated with network resources utilized by the user devices to access the content.