Method and system for improving targeted data delivery

• US 7,945,585 B1
• Filed: 10/13/2005
• Issued: 05/17/2011
• Est. Priority Date: 10/13/2005
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Tyler, Joshua R., Sorkin, Stephen P., Lukose, Rajan M.
• Primary Examiner(s): Ali; Mohammad
• Assistant Examiner(s): BUI, THUY T
• Application Number:
  US11/250,064
• Time in Patent Office: 2,042 Days
• Field of Search: 708/5, 705/27, 705/14, 705/5, 705/26, 705/37, 707/3, 707/2, 707/784, 707/731, 707/748
• US Class Current: 707/784
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A method and system for improving targeted data delivery. One method accesses a user profile based on information about a user from at least one electronic device from a set of electronic devices that is associated with the user. A targeted data offer is received, wherein the targeted data offer defines selection criteria for targeted user profiles. A determination is made as to whether the user profile satisfies the selection criteria. A non-user selected relevancy criteria based on the user profile is then utilized to characterize the targeted data offer.

Claims

1. A computer implemented method for improving targeted data delivery, said method comprising:

accessing at said computer a user profile based on information about a user, said information collected from at least one electronic device associated with said user and including web sites visited, software applications used, web searches performed, and email usage;

receiving at said computer a targeted data offer soliciting delivery of targeted data to said user, wherein said targeted data offer defines selection criteria for delivery of targeted data to said user based on said user profile;

determining at said computer that at least a portion of said user profile of said user matches said selection criteria for delivery of targeted data to said user;

determining at said computer that said targeted data offer soliciting delivery of targeted data to said user does not match user selected relevancy criteria for delivery of targeted data to said user, wherein said user selected relevancy criteria is a constraint set by said user on delivery of targeted data to said user; and

after determining that at least a portion of said user profile of said user matches said selection criteria for delivery of targeted data to said user and after determining that said targeted data offer soliciting delivery of targeted data to said user does not match said user selected relevancy criteria for delivery of targeted data to said user, allowing reception of said targeted data offer soliciting delivery of targeted data to said user if non-user selected relevancy criteria based on action of said user matches said selection criteria for delivery of targeted data to said user even though said targeted data offer soliciting delivery of targeted data to said user does not match said user selected relevancy criteria for delivery of targeted data to said user, wherein said allowing reception of said targeted data offer comprises;

utilizing at said computer at least one machine learning technique of reinforcement learning to update a learned scoring function for scoring targeted data offers; and

presenting targeted data offers to said user in a scored and ordered format.

2. The computer implemented method of claim 1, wherein said allowing reception of said targeted data offer comprises:

determining at said computer that an attribute of said targeted data offer is relevant to said user based on said user profile.

3. The computer implemented method of claim 2, wherein said allowing reception of said targeted data offer comprises:

providing said targeted data offer to said user if said targeted data offer is relevant to said user based on said user profile even though said targeted data offer does not match said user selected relevancy criteria.

4. The computer implemented method of claim 1, wherein said at least one machine learning technique comprises:

tracking at said computer each said targeted data offer said user deletes, and giving higher scores to offers dissimilar to offers said user deletes.

5. The computer implemented method of claim 1, wherein said at least one machine learning technique comprises:

tracking at said computer each said targeted data offer said user views, and giving higher scores to offers similar to offers said user views.

6. The method computer implemented of claim 1, wherein said determining that said targeted data offer does not match said user selected relevancy criteria comprises:

determining at said computer that a bid price in said targeted data offer is below an ask price defined by said user in said user selected relevancy criteria.

7. The computer implemented method of claim 1, wherein said allowing reception of said targeted data offer comprises:

superseding said user selected relevancy criteria when said non-user selected relevancy criteria determines said targeted data offer is relevant to said user profile.

8. The computer implemented method of claim 1, wherein said allowing reception of said targeted data offer comprises:

adjusting at said computer a threshold of said user selected relevancy criteria such that a number of targeted data offers provided to said user satisfies a quota of targeted data offers per time period as defined by said user in said user selected relevancy criteria.

9. The computer implemented method of claim 1, wherein said non-user selected relevancy criteria determines an order of allowing said targeted data offer utilizing criteria selected from the group including:

compensation amount, message topic, message creator, message type and message metadata.

10. The computer implemented method of claim 1, wherein said at least one machine learning technique comprises:

tracking at said computer each said targeted data offer said user accepts, and giving higher scores to offers similar to offers said user accepts.

11. The computer implemented method of claim 1, wherein said at least one machine learning technique comprises:

tracking at said computer each said targeted data offer said user marks irrelevant, and giving higher scores to offers dissimilar to offers said user marks irrelevant.

12. A computer-usable medium having computer-readable program code embodied therein for causing a computer system to perform a method for improving targeted data delivery, said method comprising:

collecting user information about a user, said user information collected from at least one electronic device associated with said user and including web sites visited, software applications used, web searches performed, and email usage;

generating a user profile based on said user information;

receiving an offer soliciting delivery of targeted data to said user, wherein said offer defines selection criteria for delivery of targeted data to said user based on said user profile;

determining that at least a portion of said user profile of said user matches said selection criteria for delivery of targeted data to said user;

determining that said targeted data offer soliciting delivery of targeted data to said user does not match user selected relevancy criteria for delivery of targeted data to said user, wherein said user selected relevancy criteria is a constraint set by said user on delivery of targeted data to said user; and

after determining that at least a portion of said user profile of said user matches said selection criteria for delivery of targeted data to said user and after determining that said targeted data offer soliciting delivery of targeted data to said user does not match said user selected relevancy criteria for delivery of targeted data to said user, allowing reception of said targeted data offer soliciting delivery of targeted data to said user if non-user selected relevancy criteria based on action of said user matches said selection criteria for delivery of targeted data to said user regardless of whether said targeted data offer soliciting delivery of targeted data to said user matches said user selected relevancy criteria for delivery of targeted data to said user, wherein said allowing reception of said targeted data offer comprises;

utilizing at least one machine learning technique of reinforcement learning to update a learned scoring function for scoring targeted data offers; and

presenting targeted data offers to said user in a scored and ordered format.

13. The computer-usable medium of claim 12, wherein said non-user selected relevancy criteria based on action of said user comprises:

tracking which targeted data offers said user accepts.

14. The computer-usable medium of claim 12, wherein said non-user selected relevancy criteria based on action of said user comprises:

tracking which targeted data offers said user marks irrelevant.

15. The computer-usable medium of claim 12, wherein said non-user selected relevancy criteria based on action of said user comprises:

tracking which targeted data offers said user deletes.

16. The computer-usable medium of claim 12, wherein said non-user selected relevancy criteria based on action of said user comprises:

tracking which targeted data offers said user views.

17. The computer-usable medium of claim 12, wherein said determining that said targeted data offer does not match said user selected relevancy criteria comprises:

determining that a bid price in said targeted data offer is below an ask price defined by said user in said user selected relevancy criteria.

18. The computer-usable medium of claim 12, wherein said allowing reception of said targeted data offer comprises:

superseding the user selected relevancy criteria when said non-user selected relevancy criteria determines said targeted data offer is relevant to said user profile.

19. The computer-usable medium of claim 12, wherein said allowing reception of said targeted data offer comprises:

adjusting a threshold of said user selected relevancy criteria such that a number of targeted data offers provided to said user satisfies a total number of offers per time period as defined by said user in said user selected relevancy criteria.