Targeted advertisement placement based on explicit and implicit criteria matching

• US 8,296,179 B1
• Filed: 04/25/2008
• Issued: 10/23/2012
• Est. Priority Date: 05/02/2007
• Status: Active Grant

• Current Assignee: Monster Worldwide Incorporated (Randstad NV)
• Original Assignee: Monster Worldwide Incorporated (Randstad NV)
• Inventors: Rennison, Earl
• Primary Examiner(s): Kyle, Charles
• Assistant Examiner(s): Madamba, Clifford
• Application Number:
  US12/110,213
• Time in Patent Office: 1,642 Days
• Field of Search: 705/14
• US Class Current: 705/14
• CPC Class Codes: G06Q 30/0251 Targeted advertisements

Abstract

A targeted advertisement placement system allows an advertiser to place an ad in front of an audience based on specific interests of the users in that audience, and based on the advertiser'"'"'s own specific interests in users who should view the ad. The ad placement system collects information about a population of users and a population of advertisers to create profiles that can include both explicit and implicit criteria defining the user'"'"'s/advertiser'"'"'s interest. The explicit criteria can be explicitly provided to the ad placement system, can be extracted from documents provided by the user and/or the advertiser, etc. The implicit criteria include information learned by the system about the interests of both advertisers and users based on preferences implicitly expressed when providing feedback on search results. Using these criteria collected from advertisers and users to build profiles, the system can compare the profiles for matches to determine the best audience before which to place an advertisement.

Claims

1. A method for targeting advertisements to users, the method comprising:

using at least one computer system to perform method steps comprising;

for a plurality of advertisers;

receiving explicit criteria from a plurality of advertisers;

providing search results to each of the plurality of advertisers responsive to the explicit criteria;

receiving feedback from each of the plurality of advertisers about the search results;

determining implicit criteria about which users the plurality of advertisers wish to target based on the feedback received from the plurality of advertisers, the implicit criteria indicating hidden preferences of the advertisers identified from the feedback from each of the plurality of advertisers;

for a plurality of users;

receiving explicit user criteria providing information about each of the users, the criteria providing information about which advertisements the users should view;

providing the users with search results using the explicit user criteria for one or more searches conducted by each of the users;

receiving user feedback regarding the search results;

determining implicit user criteria based on the feedback provided by the users;

creating a plurality of advertiser/ad profiles each using the explicit and implicit advertiser/ad criteria for a given advertiser/ad;

creating a plurality of user profiles each using the explicit and implicit user criteria for a given user;

comparing the advertiser/ad profiles and the user profiles to generate relevance scores and determine matches between a particular advertiser/ad profile and a particular user profile;

determining which of the advertisements to present to each of the plurality of users based on placement factors including computing relevance of each of the advertisements to each of the plurality of users, computing value of each of the plurality of users to each of the advertisers based on the relevance of each of the advertisements to each of the plurality of users, and computing a price for presenting the advertisement to each of the plurality of users; and

presenting to each of the plurality of users the advertisements for which a positive determination to place the advertisement is made.

2. The method of claim 1, wherein the plurality of advertisers are job providers presenting advertisements to job candidates and the plurality of users are job candidates conducting searches for job openings.

3. The method of claim 2, wherein the advertiser/ad profiles indicate preferences of the job providers in job candidates to fill the job opening and the user profiles indicate preferences of the job candidates in jobs.

4. The method of claim 1, wherein the plurality of advertisers are job candidates placing an advertisement before a job provider and the plurality of users are job providers conducting a search for job candidates to fill job openings.

5. The method of claim 4, wherein the advertiser/ad profiles indicate preferences of the job candidates in jobs and the user profiles indicate preferences of the job providers in job candidates to fill the job openings.

6. The method of claim 1, further comprising:

extracting concepts from information provided by the plurality of users and the plurality of advertisers; and

comparing the concepts extracted to a knowledge base of linked concepts to generate a set of relevant concepts for use in matching the user profiles with the advertiser/ad profiles.

7. The method of claim 1, wherein one or more of the user criteria for a user are explicit criteria extracted from documents provided by the each of the plurality of users.

8. The method of claim 1, wherein one or more of the advertiser criteria are explicit criteria extracted from documents provided by the advertisers.

9. The method of claim 1, wherein one or more of the user criteria are implicit criteria learned based on feedback provided by the users regarding search results received in one or more searches conducted.

10. The method of claim 1, wherein determining which of the advertisements to present further comprises determining an order in which to present the advertisements to the users.

11. The method of claim 1, wherein determining which of the advertisements to present further comprises determining a location at which the advertisement should be placed relative to other advertisements being presented to the user.

12. The method of claim 1, wherein the relevance measure comprises a relevance matrix providing the relevance scores for the relevance of a given advertisement to each of the users.

13. The method of claim 12, wherein the value measure comprises a value matrix providing value scores for the value of presenting a given advertisement to each of the users, the value scores calculated based on the relevance scores.

14. The method of claim 13, wherein the value is calculated in terms of supply and demand, where supply equals a sum of the relevance scores of the plurality of users against each advertisement, where demand equals a sum of the relevance scores of each advertisement against the users, and where supply and demand are inversely related.

15. The method of claim 14, wherein the sum for supply is divided by the total number of users and the sum for demand is divided by the total number of advertisements, and wherein the value equals the demand multiplied by the inverse of the supply.

16. The method of claim 14, wherein the value is equal to the square root of the log of the sum of the relevance scores of the users for all of the users divided by the log of the number of users times the log of the number of advertisements divided by the sum of the relevance scores of the users for all of the ads divided by the log of the number of ads.

17. The method of claim 15, wherein the price of each advertisement is computed by multiplying a maximum price that the advertiser is willing to pay for the advertisement by the value of the advertisement for the given user, where the value of the advertisement for the user is stored in a value matrix, thereby forming a vector of prices for each of the advertisements for the given user, the vector being sorted in descending order with higher ranking advertisements being presented first.

18. The method of claim 17, wherein the price is variable and is determined as a price value vector by taking the price that the advertiser is willing to pay to present the advertisement to a given user multiplied by the value for that user.

19. The method of claim 17, wherein the price is a fixed price that the advertiser is willing to pay for presenting the advertisement to any of the users.

20. The method of claim 1, further comprising accepting one or more price bids from advertisers to present their advertisements to the users according to a quality of the match between the profiles.

21. The method of claim 1, further comprising setting a value threshold indicating that the advertiser is unwilling to pay to present the advertisement to users having values below the threshold.

22. The method of claim 1, wherein the advertisement is presented to the user via a method selected from a group consisting of:

placing the advertisement amongst search results obtained in a search, placing the advertisement when the user logs in to an account, and sending the advertisement to the user.

23. The method of claim 1, wherein one or more of the steps is performed offline.

24. The method of claim 1, wherein one or more of the steps is performed online.

25. The method of claim 1, further comprising:

receiving a search string provided by the user conducting a search of a database;

computing a search relevance of the advertisement to the user based on the search string; and

determining whether to present the advertisement to the user based the results of each of the computations.

26. The method of claim 25, further comprising rotating an advertisement inventory so the user is presented with different advertisements over time.

27. The method of claim 25, wherein the steps of receiving a search string, computing a search relevance, determining whether to present the advertisement, and presenting the advertisement are performed online, and the remaining steps are performed offline, before the user conducts the search.

28. The method of claim 1, wherein computing the price further comprises:

determining the price that the advertiser is willing to pay to present the advertisement to the user; and

sorting the price relative to prices for other advertisements, the prices sorted in descending order with higher ranking advertisements being presented first.

29. The method of claim 28, wherein the search relevance is computed by determining a search relevance score for the relevance of the advertisement to the user based on the search string.

30. The method of claim 29, wherein the search relevance score is combined with information regarding the price computed for determining whether to present the advertisement to the user, and where or when to present the advertisement relative to other advertisements.

31. The method of claim 1, wherein a user has a higher value when fewer total users match the advertiser/ad criteria or when more advertisers want to advertise to the user.

32. A system for targeting advertisements to users, the system comprising:

at least one computer system;

a computer-readable storage medium storing software modules for execution by the at least one computer system, the modules comprising;

a criteria collection module for;

receiving explicit criteria from a plurality of advertisers;

providing search results to each of the plurality of advertisers responsive to the explicit criteria;

receiving feedback from each of the plurality of advertisers about the search results;

determining implicit criteria about which users the plurality of advertisers wish to target based on the feedback received from the plurality of advertisers, the implicit criteria indicating hidden preferences of the advertisers identified from the feedback from each of the plurality of advertisers;

the criteria collection module for;

receiving explicit user criteria providing information about each of the users, the criteria providing information about which advertisements the users should view;

providing the users with search results using the explicit user criteria for one or more searches conducted by each of the users;

receiving user feedback regarding the search results;

determining implicit user criteria based on the feedback provided by the users;

a profile creation module for creating a plurality of advertiser/ad profiles each using the explicit and implicit advertiser/ad criteria for a given advertiser/ad;

the profile creation module for creating a plurality of user profiles each using the explicit and implicit user criteria for a given user;

a profile comparison module for comparing the advertiser/ad profiles and the user profiles to generate relevance scores and determine matches between a particular advertiser/ad profile and a particular user profile;

an ad targeting manager for determining which of the advertisements to present to each of the plurality of users based on placement factors including computing relevance of each of the advertisements to each of the plurality of users, computing value of each of the plurality of users to each of the plurality of advertisers based on the relevance of each of the advertisements to each of the plurality of users, and computing a price for presenting the advertisement to each of the plurality of users; and

the ad targeting manager for presenting to each of the plurality of users the advertisements for which a positive determination to place the advertisement is made.

33. The system of claim 32, wherein the advertisers are job providers interested in placing advertisements before job candidates and the users are job candidates interested in conducting searches for job openings, the advertiser/ad profiles each indicating preferences of different job providers in job candidates to fill a job opening and the user profiles indicating preferences of the job candidates in a job.

34. The system of claim 32, wherein the plurality of advertisers are job candidates interested in placing advertisements before a job provider and the plurality of users are job providers interested in conducting searches for job candidates to fill a job opening, the advertiser/ad profiles indicating preferences of the job candidates in a job and the user profiles indicating preferences of the job providers in a job candidate to fill the job opening.

35. The system of claim 32, further comprising:

a concept extraction module for extracting concepts from information provided by the plurality of users and the plurality of advertisers; and

the concept extraction module for comparing the concepts extracted to one or more knowledge bases of linked concepts to generate a set of relevant concepts for use in matching the user profiles with the advertiser/ad profiles.

36. The system of claim 32, wherein one or more of the user criteria are explicit criteria extracted from documents provided by the users, and wherein one or more of the advertiser criteria are explicit criteria extracted from documents provided by the advertisers.

37. The system of claim 32, wherein one or more of the user criteria are implicit criteria learned by the system based on feedback provided by the users regarding search results received in one or more searches conducted, and wherein one or more of the advertiser criteria are implicit criteria learned by the system based on feedback provided by the advertisers regarding search results received in one or more searches conducted.

38. The system of claim 32, wherein the ad targeting manager is further configured for determining an order in which to present the advertisements to the users and determining a location at which the advertisement should be placed relative to other advertisements being presented to the user.

39. The system of claim 32, wherein the ad targeting manager is further configured for computing the relevance measure comprising a relevance matrix providing relevance scores for the relevance of a given advertisement to each of the users.

40. The system of claim 39, wherein the ad targeting manager is further configured for computing the value measure comprising a value matrix providing value scores for the value of presenting a given advertisement to each of the users, the value scores calculated based on the relevance scores.

41. The system of claim 40, wherein the ad targeting manager is further configured for calculating the value in terms of supply and demand, where supply equals a sum of the relevance scores of the users against each advertisement, where demand equals a sum of the relevance scores of each advertisement against the users, wherein the sum for supply is divided by the total number of users and the sum for demand is divided by the total number of advertisements and wherein supply and demand are inversely related.

42. The system of claim 40, wherein the value is equal to the square root of the log of the sum of the relevance scores of the users for all of the users divided by the log of the number of users times the log of the number of advertisements divided by the sum of the relevance scores of the users for all of the ads divided by the log of the number of ads.

43. The system of claim 39, wherein the ad targeting manager is further configured for:

computing the price of each advertisement by multiplying a maximum price that the advertiser is willing to pay for the advertisement by the value of the advertisement for the given user, where the value of the advertisement for the user is stored in a value matrix, thereby forming a vector of prices for each of the advertisements for the given user; and

sorting the vector in descending order with higher ranking advertisements being presented first.

44. The system of claim 43, wherein the price is variable and the ad targeting manager is configured for determining the price as a price value vector, the price that the advertiser is willing to pay to present the advertisement to a given user being multiplied by the value for that user.

45. The system of claim 43, wherein the price is a fixed price that the advertiser is willing to pay for presenting the advertisement to any of the users.

46. The system of claim 32, wherein the ad targeting module is further configured for accepting one or more price bids from advertisers to present their advertisements to the users according to a quality of the match between the profiles.

47. A method for targeting advertisements to users, comprising:

using at least one computer system to perform steps comprising;

receiving explicit criteria from an advertiser about preferences for users to be presented with an advertisement;

providing search results to the advertiser responsive to the explicit criteria;

receiving feedback from the advertiser about the search results;

determining implicit criteria about which users the advertiser wishes to target based on the feedback received from the advertiser, the implicit criteria indicating hidden preferences of the advertiser identified from the feedback from the advertiser;

for a plurality of users;

receiving explicit user criteria providing information about each of the users, the criteria providing information about which advertisements the users should view;

providing the users with search results using the explicit user criteria for one or more searches conducted by each of the users;

receiving user feedback regarding the search results;

determining implicit user criteria based on the feedback provided by the users;

creating an advertiser profile using the explicit and implicit criteria for the advertiser;

creating a plurality of user profiles, each using explicit and implicit user criteria for a given user;

comparing the advertiser profile and each of the user profiles to generate relevance scores and determine matches between the advertiser profile and a particular user profile;

determining which of the advertisements to present to each of the plurality of users based on placement factors including;

computing relevance of each of the advertisements to each of the plurality of users, computing value of each of the plurality of users to the advertiser based on the relevance of each of the advertisements to each of the plurality of users, and computing a price for presenting the advertisement to each of the plurality of users; and

presenting to each of the plurality of users the advertisements for which a positive determination to place the advertisement is made.

48. The method of claim 47, wherein the advertiser is a job provider presenting advertisements to job candidates and the plurality of users are job candidates conducting searches for job openings.

49. The method of claim 48, wherein the explicit or implicit criteria of a user include interests or preferences in a job opening.

50. The method of claim 48, wherein the explicit or implicit criteria of an advertiser includes preferences for an employee.

51. The method of claim 48, wherein the explicit or implicit criteria for a user is collected from the resume of the user.

52. The method of claim 48, wherein the explicit criteria of the user is collected from a query directed to the user, a questionnaire, a registration form, or a user'"'"'s search query.

53. The method of claim 48, wherein the explicit criteria of the advertiser is collected from a query directed to the user, a questionnaire, a registration form, or a search query.

54. The method of claim 48, wherein the feedback from the advertiser is received in the form of a vote, a rating, or information provided by the advertiser.