Ad placement

• US 10,217,128 B2
• Filed: 09/12/2012
• Issued: 02/26/2019
• Est. Priority Date: 11/08/1999
• Status: Expired due to Term

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Inventors: Ferber, John B., Ferber, Scott, Kretsinger, Stein E., Luenberger, Robert, Luenberger, David
• Primary Examiner(s): Munson, Patricia H
• Application Number:
  US13/612,631
• Publication Number:
  US 20130054347A1
• Time in Patent Office: 2,358 Days
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0241 Advertisements
• CPC Class Codes: G06Q 30/0244 Optimization
• CPC Class Codes: G06Q 30/0246 Traffic
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 30/0254 based on statistics
• CPC Class Codes: G06Q 30/0269 based on user profile or attribute
• CPC Class Codes: G06Q 30/0277 Online advertisement

Abstract

This invention concerns optimal ad selection for Web pages by selecting and updating an attribute set, obtaining and updating an ad-attribute profile, and optimally choosing the next ad. The present invention associates a set of attributes with each customer. The attributes reflect the customers'"'"' interests and they incorporate the characteristics that impact ad selection. Similarly, the present invention associates with each ad an ad-attribute profile in order to calculate a customer'"'"'s estimated ad selection probability and measure the uncertainty in that estimate. An ad selection algorithm optimally selects which ad to show based on the click probability estimates and the uncertainties regarding these estimates.

Claims

1. A method comprising:

estimating, by at least one processor, principal component vectors for each advertisement of a plurality of advertisements served via one or more network-based mediums based on a first heuristic;

determining, by the at least one processor, a click probability for each advertisement of the plurality of advertisements served via the one or more network-based mediums by employing the estimated principal component vectors and a second heuristic, wherein employing the estimated principal component vectors causes the at least one processor to determine the click probability for each advertisement served via the one or more network-based mediums at a given level of accuracy using a reduced number of impressions, and wherein the second heuristic differs from the first heuristic; and

serving one or more advertisements selected from the plurality of advertisements over a communications network, to the one or more network-based mediums, based on the determined click probabilities.

2. The method as recited in claim 1, wherein the second heuristic comprises a least squares regression analysis, and wherein employing the estimated principal component vectors further causes the at least one processor to quickly and efficiently determine the click probability for each advertisement served via the one or more network-based mediums.

3. The method as recited in claim 2, further comprising determining an estimated selection probability for each advertisement from the plurality of advertisements, the estimated selection probability indicating a likelihood that a user will select a given advertisement from the plurality of advertisements, and wherein determining the click probability for each advertisement from the plurality of advertisements is further based on the estimated selection probability.

4. The method as recited in claim 2, further comprising maintaining a user profile, the user profile including user attributes reflecting interests of a user.

5. The method as recited in claim 4, further comprising updating the user attributes of the user profile based on one or more of Internet sites visited by the user, advertisements selected by the user, or searching by the user.

6. The method as recited in claim 5, further comprising maintaining an advertisement profile for the one or more advertisements of the plurality of advertisements, each advertisement profile including ad-attributes that reflect how much an advertisement from the plurality of advertisements correlates to a given user attribute.

7. The method as recited in claim 6, wherein the determined click probability for a particular advertisement from the plurality of advertisements is a function of consumer attributes of a given user and the ad-attributes for the particular advertisement from the plurality of advertisements.

8. The method as recited in claim 2, wherein the second heuristic comprising a least squares regression analysis comprises an ordinary least squares regression or a generalized least squares regression.

9. The method as recited in claim 1, further comprising serving the one or more advertisements from the plurality of advertisements to a marketing medium of the one or more network-based mediums.

10. The method as recited in claim 9, wherein the marketing medium is on a mobile device.

11. The method as recited in claim 9, wherein the marketing medium comprises a software program.

12. The method as recited in claim 9, wherein the marketing medium comprises an Internet browser.

43. The method of claim 1, wherein:

the first heuristic comprises a click probability heuristic; and

second heuristic is a heuristic other than a click probability heuristic.

47. The method of claim 1, wherein the second heuristic comprises multiple heuristics.

13. A non-transitory computer-readable storage medium including a set of instructions that, when executed, cause at least one processor to perform steps comprising:

estimating principal component vectors for each advertisement of a plurality of advertisements served via one or more network-based mediums based on a first heuristic;

determining a click probability for each advertisement of the plurality of advertisements served via the one or more network-based mediums by employing the estimated principal component vectors and a second heuristic, wherein employing the estimated principal component vectors causes the at least one processor to determine the click probability for each advertisement served via the one or more network-based mediums at a given level of accuracy using a reduced number of impressions, and wherein the second heuristic differs from the first heuristic; and

serving one or more advertisements selected from the plurality of advertisements over a communications network, to the one or more network-based mediums, based on the determined click probabilities.

14. The computer-readable storage medium as recited in claim 13, wherein employing the estimated principal component vectors further causes the at least one processor to quickly and efficiently determine the click probability for each advertisement served via the one or more network-based mediums.

15. The computer-readable storage medium as recited in claim 13, further comprising instructions that, when executed, cause the at least one processor to determine an estimated selection probability for each advertisement form the plurality of advertisements, the estimated selection probability indicating a likelihood that a user will select a given advertisement from the plurality of advertisements, and wherein determining the click probability for each advertisement from the plurality of advertisements is further based on the estimated selection probability.

16. The computer-readable storage medium as recited in claim 14, further comprising instructions that, when executed, cause the at least one processor to maintain a user profile, the user profile including user attributes reflecting interests of a user.

17. The computer-readable storage medium as recited in claim 16, further comprising instructions that, when executed, cause the at least one processor to update the user attributes of the user profile based on one or more of Internet sites visited by the user, advertisements selected by the user, or searching by the user.

18. The computer-readable storage medium as recited in claim 17, further comprising instructions that, when executed, cause the at least one processor to maintain an advertisement profile for the one or more advertisements of the plurality of advertisements, each advertisement profile including ad-attributes that reflect how much an advertisement from the plurality of advertisements correlates to a given user attribute.

19. The computer-readable storage medium as recited in claim 18, wherein the determined click probability for a particular advertisement from the plurality of advertisements is a function of consumer attributes of a given user and the ad-attributes for the particular advertisement from the plurality of advertisements.

20. The computer-readable storage medium as recited in claim 13, wherein the second heuristic comprises a least squares regression analysis.

21. The computer-readable storage medium as recited in claim 13, further comprising instructions that, when executed, cause the at least one processor to serve the one or more advertisements from the plurality of advertisements to a marketing medium of the one or more network-based mediums.

22. The computer-readable storage medium as recited in claim 21, wherein the marketing medium is on a mobile device.

23. The computer-readable storage medium as recited in claim 21, wherein the marketing medium comprises a software program.

24. The computer-readable storage medium as recited in claim 21, wherein the marketing medium comprises an Internet browser.

44. The computer-readable storage medium as recited in claim 13, wherein:

the first heuristic comprises a click probability heuristic; and

the second heuristic is a heuristic other than a click probability heuristic.

25. A method comprising:

determining, by at least one processor, an estimated selection probability for one or more advertisements of a plurality of advertisements served via one or more network-based mediums, wherein;

the estimated selection probability indicates a likelihood that a user will select a given advertisement served via the one or more network-based mediums;

determining the estimated selection probability comprises by employing a principal component analysis; and

determining the estimated selection probability comprises using least squares regression based on the principal component analysis such that the at least one processor determines the estimated selection probability for each advertisement served via the one or more network-based mediums at a given level of accuracy using a reduced number of impressions; and

serving an advertisement to the one or more network-based mediums with a high estimated selection probability.

26. The method as recited in claim 25, wherein determining the estimated selection probability further comprises calculating a click-thru-rate for the one or more advertisements of the plurality of advertisements.

27. The method as recited in claim 25, further comprising binomially updating the estimated selection probability of the one or more advertisements of the plurality of advertisements.

28. The method as recited in claim 27, further comprising:

maintaining a user profile, the user profile including user attributes reflecting interests of the user;

maintaining an advertisement profile for the one or more advertisements of the plurality of advertisements, each advertisement profile indicating how much a given advertisement correlates to a given user attribute; and

wherein the estimated selection probability for a particular advertisement is a function of consumer attributes of a given user and the advertisement profile for the particular advertisement.

29. The method as recited in claim 25, wherein serving an advertisement with a high estimated selection probability comprises serving the advertisement with the highest estimated selection probability.

30. The method as recited in claim 25, further comprising serving the advertisement with a high estimated selection probability to a marketing medium of the one or more network-based mediums.

31. The method as recited in claim 30, wherein the marketing medium is on a mobile device.

32. The method as recited in claim 31, wherein the marketing medium comprises a software program.

33. The method as recited in claim 25, wherein using the principal analysis causes the at least one processor to determine the estimated selection probability for the one or more advertisements of the plurality of advertisements served via the one or more network-based mediums at a given level of accuracy using a reduced number of impressions.

46. The method of claim 25, wherein the principal component analysis comprises:

estimating a principal component; and

estimating a click probability based on the principal component.

34. A method comprising:

estimating, by at least one processor, principal component vectors for each advertisement of an advertisement type served via one or more network-based mediums based on a first heuristic;

determining, by the at least one processor, a click probability for each advertisement of the advertisement type served via the one or more network-based mediums by employing the estimated principal component vectors and a second heuristic, wherein employing the estimated principal component vectors causes the at least one processor to determine the click probability for each advertisement served via the one or more network-based mediums at a given level of accuracy using a reduced number of impressions, and wherein the second heuristic differs from the first heuristic; and

serving one or more advertisements of the advertisement type to one or more mobile devices on the one or more network-based mediums based on the determined click probabilities.

35. The method as recited in claim 34, wherein employing the estimated principal component vectors further causes the at least one processor to quickly and efficiently determine the click probability for each advertisement served via the one or more network-based mediums.

36. The method as recited in claim 34, further comprising determining an estimated selection probability for each advertisement of the advertisement type, the estimated selection probability indicating a likelihood that a user will select a given advertisement of the advertisement type, and wherein determining the click probability for each advertisement of the advertisement type is further based on the estimated selection probability.

37. The method as recited in claim 35, further comprising maintaining a user profile, the user profile including user attributes reflecting interests of a user.

38. The method as recited in claim 37, further comprising updating the user attributes of the user profile based on one or more of Internet sites visited by the user, advertisements selected by the user, searching by the user, or activities of the user on a marketing medium.

39. The method as recited in claim 38, further comprising maintaining an advertisement profile for the one or more advertisements of the advertisement type, each advertisement profile including ad-attributes that reflect how much an advertisement correlates to a given user attribute.

40. The method as recited in claim 39, wherein the determined click probability for a particular advertisement of the advertisement type is a function of consumer attributes of a given user and the ad-attributes for the particular advertisement.

41. The method as recited in claim 34, wherein the second heuristic comprises a least squares regression analysis.

42. The method as recited in claim 34, further comprising serving the one or more advertisements of the advertisement type to a marketing medium of the one or more network-based mediums.

45. The method of claim 34, wherein:

the first heuristic comprises a click probability heuristic; and

the second heuristic is a heuristic other than a click probability heuristic.