Client-centered usage classification

• US 8,166,104 B2
• Filed: 03/19/2009
• Issued: 04/24/2012
• Est. Priority Date: 03/19/2009
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Shein, Barak, Karmon, Kfir, Yariv, Eran, Varshavsky, Roy
• Primary Examiner(s): HIGA, BRENDAN Y
• Application Number:
  US12/406,964
• Publication Number:
  US 20100241687A1
• Time in Patent Office: 1,132 Days
• Field of Search: 709223-224, 709201-203, 726/26, 705 1449- 1467, 707/754
• US Class Current: 709/203
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 30/0255 based on user history

Abstract

User input actions from one or more user input devices can be received, and unfiltered user input information about the user input actions can be collected at a client computing environment. The unfiltered user input information can be filtered to produce filtered user input information. The filtered user input information can exclude personal information and include non-personal information. In addition, a user profile can be created using the filtered user input information.

Claims

1. One or more computer-readable storage media not consisting of a propagated signal, the one or more computer-readable storage media having computer-executable instructions embodied thereon that, when executed, perform acts comprising:

receiving user input actions from one or more user input devices at a client computing environment;

collecting unfiltered user input information about the user input actions at the client computing environment;

filtering the unfiltered user input information to produce filtered user input information at the client computing environment, the filtered user input information excluding personal information of the unfiltered user input information and including non-personal information of the unfiltered user input information, filtering comprising excluding one or more personal patterns of user action representations, each personal pattern comprising user action representations that reveal personal information when taken together in the pattern; and

creating a user profile using the filtered user input information.

2. The one or more computer-readable media of claim 1, wherein filtering comprises limiting user input received from one or more user input devices in the client computing environment.

3. The one or more computer-readable media of claim 1, wherein filtering comprises excluding from the filtered user input information one or more personal user action representations that each reveal personal information without reference to other user action representations.

4. The one or more computer-readable media of claim 1, wherein creating a user profile comprises classifying the filtered user input information according to one or more segment rules, and wherein the user profile comprises one or more correlations of the filtered user input information to one or more segments according to the one or more segment rules.

5. The one or more computer-readable media of claim 1, wherein the user profile comprises one or more assignments to one or more segments.

6. The one or more computer-readable media of claim 1, wherein the user profile comprises one or more representations of one or more probabilities of the filtered user input information correlating to one or more segments.

7. The one or more computer-readable media of claim 1, wherein one or more of the user input actions results in information being sent over a global computer network.

8. The one or more computer-readable media of claim 1, wherein creating a user profile comprises classifying the filtered user input information according to one or more tentative segments, and the acts further comprise using the user profile to determine whether to include segment rules corresponding to the one or more tentative segments in a set of segment rules.

9. The one or more computer-readable media of claim 1, wherein the acts further comprise receiving user segment rules from a server computing environment over a global computer network, and wherein creating a user profile comprises classifying the filtered user input information at the client computing environment according to the user segment rules.

10. The one or more computer-readable media of claim 1, wherein the acts further comprise sending the user profile to a server computing environment over a global computer network.

11. The one or more computer-readable media of claim 10, wherein the acts further comprise obfuscating the user profile to generate an obfuscated user profile, and wherein sending the user profile comprises sending the obfuscated user profile.

12. The one or more computer-readable media of claim 1, wherein:

filtering comprises filtering according to one or more sensitivity rules that define personal information to be excluded from use in creating the user profile;

the user profile comprises one or more correlations to one or more segments according to one or more segment rules; and

the acts further comprise;

sending messages over a global computer network in response to the user input actions;

sending the user profile to a remote server over the global computer network; and

receiving one or more advertisements over the global computer network in response to sending the user profile.

13. A computer-implemented method, comprising:

collecting user input information at a client computing environment, the user input information representing user input actions received from one or more user input devices in the client computing environment;

storing the user input information in one or more storage devices in the client computing environment;

filtering the user input information to produce filtered user input information at the client computing environment, the filtered user input information excluding personal information of the unfiltered user input information and including non-personal information of the unfiltered user input information;

classifying the collected user input information at the client computing environment to create an initial user profile;

obfuscating the initial user profile to create an obfuscated user profile, the initial user profile comprising one or more genuine correlations to one or more segments to which the user input information has been correlated when classifying, and the obfuscated user profile comprising the one or more genuine correlations and one or more fake correlations to one or more segments to which the user input information has not been correlated when classifying; and

sending the obfuscated user profile to a server computing environment.

14. The method of claim 13, further comprising:

receiving in response to the obfuscated user profile one or more units of information corresponding to the one or more fake correlations and one or more units of information corresponding to the one or more genuine correlations; and

displaying on a display in the client computing environment one or more representations of the one or more units of information corresponding to the one or more genuine correlations, but not displaying representations of the one or more units of information corresponding to the one or more fake correlations.

15. The method of claim 13, the user profile including one or more representations of one or more probabilities of the collected user input information correlating to one or more segments.

16. A computer system comprising:

one or more storage devices storing user input information representing user input actions received from one or more input devices in the computer system;

at least one processor; and

memory comprising instructions stored thereon that when executed by at least one processor cause at least one processor to perform acts comprising;

classifying the stored user input information; and

excluding personal user input information from being used in classifying the stored user input information, excluding the personal user input information from being used in classifying comprising excluding use of patterns of representations of user actions that when taken together reveal personal user input information.

17. The computer system of claim 16, wherein excluding the personal user input information from being used in classifying comprises generalizing at least a portion of the personal user input information.

18. The computer system of claim 16, wherein classifying comprises applying segment rules to the stored user input information.