USING CONCEPTS FOR AD TARGETING

• US 20100287056A1
• Filed: 07/16/2010
• Published: 11/11/2010
• Est. Priority Date: 11/24/2003
• Status: Abandoned Application

• Current Assignee: Google Inc. (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: KONINGSTEIN, Ross, Shazeer, Noam, Harik, Georges R., Spitkovsky, Valentin
• Application Number:
  US12/837,883
• Publication Number:
  US 20100287056A1
• Time in Patent Office: Days
• US Class Current: 705/14.73
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0243 Comparative campaigns
• CPC Class Codes: G06Q 30/0277 Online advertisement

Abstract

Concept similarity may be used to help resolve ambiguities with respect to ads served using, at least, keyword targeting. More specifically, concept similarity may be used to help determine ad relevancy and/or ad scores.

Claims

1. A computer-implemented method comprising:

a) determining, by an ad serving system including at least one computer on a network, at least one candidate concept of an ad accepted from an advertiser, each of the at least one candidate concepts determined having an ad concept score;

b) presenting, by the ad serving system, the determined at least one candidate concept of the ad to the advertiser;

c) accepting, by the ad serving system, feedback from the advertiser to the presented at least one candidate concept;

d) adjusting, by the ad serving system, the ad concept score associated with the at least one candidate concept using the accepted advertiser feedback;

e) storing, by the ad serving system, the adjusted ad concept score in association with the at least one candidate concept and in association with the ad;

f) determining, by the ad serving system, at least one request concept associated with a received user request, each of the at least one request concepts having an associated request concept score;

g) determining, by the ad serving system, a similarity score between the ad and the received user request using, at least, the at least one candidate concept of the ad and its adjusted ad concept score in combination with the at least one request concept and its request concept score; and

h) controlling, by the ad serving system, serving of the ad using at least the determined similarity score.

2. The computer-implemented method of claim 1 wherein the candidate concept of the ad includes context information, the context information being a representation of meaning that is determined by analyzing a sequence of at least one of (A) word searches or (B) user actions as the result of word searches.

3. The computer-implemented method of claim 1 further comprising:

determining, by the ad serving system, a further candidate concept having an associated score using advertiser feedback; and

presenting, by the ad serving system, the determined further candidate concept to the advertiser.

4. The computer-implemented method of claim 1 wherein the candidate concept is a previously processed search query to which the ad would have been relevant.

5. The computer-implemented method of claim 1 wherein the at least one request concept was determined by:

1) determining, by the ad serving system, at least one request concept having an associated request concept score using the received user request, the at least one request concept including context information, and wherein the request concept score associated with the request concept is determined based on the similarity between the determined request concept and the received user request;

2) generating, by the ad serving system, a representation of the determined at least one request concept, wherein the request concept score associated with the determined at least one request concept in the generated representation is adjusted using performance information of advertisements that have been served pursuant to the request concept; and

3) storing, by the ad serving system, the adjusted value in association with the at least one request concept.

6. The computer-implemented method of claim 5 wherein the performance information is advertisement selection information.

7. The computer-implemented method of claim 5 wherein the performance information is conversion information.

8. The computer-implemented method of claim 5 wherein the performance information of advertisements that have been served pursuant to the request concept used is tracked performance information of advertisements served pursuant to the concept.

9. The computer-implemented method of claim 8 wherein adjusting the request concept score associated with the request concept relative to the request includes increasing the request concept score associated with the request concept relative to the request if the tracked performance information is above a threshold performance level.

10. The computer-implemented method of claim 8 wherein adjusting the request concept score associated with the request concept relative to the request includes decreasing the request concept score associated with the request concept relative to the request if the tracked performance information is below a threshold performance level.

11. The computer-implemented method of claim 8 wherein adjusting the request concept score associated with the request concept relative to the request uses the tracked performance of the request concept relative to tracked performance of at least one other concept

12. The computer-implemented method of claim 8 wherein the performance information is advertisement selection information.

13. The computer-implemented method of claim 8 wherein the performance information is conversion information.

14. A computer-implemented method comprising.

a) determining, by an ad serving system including at least one computer on a network, at least one targeting concept from targeting criteria information associated with an ad, each of the at least one targeting concepts determined having a associated targeting concept score;

b) adjusting, by the ad serving system, the targeting concept score associated with the at least one targeting concept using at least information from other ads using the same or similar targeting criteria information;

c) storing, by the ad serving system, the adjusted targeting concept score in association with the at least one targeting concept and in association with the ad;

d) determining, by the ad serving system, at least one request concept associated with a received user request, each of the at least one request concepts having an associated request concept score;

e) determining, by the ad serving system, a similarity score between the ad and the received user request using, at least, the determined at least one targeting concept of the ad and its adjusted targeting concept score in combination with the at least one request concept and its request concept score; and

f) controlling, by the ad serving system, serving of the ad using at least the determined similarity score.

15. The computer-implemented method of claim 14 wherein the targeting concept includes context information, the context information being a representation of meaning that is determined by analyzing a sequence of at least one of (A) word searches or (B) user actions as the result of word searches.

16. Apparatus comprising:

a) at least one processor;

b) at least one communications interface; and

c) at least one storage device, the storage device storing program instructions which, when executed by the at least one processor, perform a method including;

1) determining at least one candidate concept of an ad accepted from an advertiser, each of the at least one candidate concepts determined having an ad concept score;

2) presenting the determined at least one candidate concept of the ad to the advertiser;

3) accepting feedback from the advertiser to the presented at least one candidate concept;

4) adjusting the ad concept score associated with the at least one candidate concept using the accepted advertiser feedback;

5) storing the adjusted ad concept score in association with the at least one candidate concept and in association with the ad;

6) determining at least one request concept associated with a received user request, each of the at least one request concepts having an associated request concept score;

7) determining a similarity score between the ad and the received user request using, at least, the at least one candidate concept of the ad and its adjusted ad concept score in combination with the at least one request concept and its request concept score; and

8) controlling serving of the ad using at least the determined similarity score.

17. Apparatus comprising:

a) at least one processor;

b) at least one communications interface; and

c) at least one storage device, the storage device storing program instructions which, when executed by the at least one processor, perform a method including;

1) determining at least one targeting concept from targeting criteria information associated with an ad, each of the at least one targeting concepts determined having a associated targeting concept score;

2) adjusting the targeting concept score associated with the at least one targeting concept using at least information from other ads using the same or similar targeting criteria information;

3) storing the adjusted targeting concept score in association with the at least one targeting concept and in association with the ad;

4) determining at least one request concept associated with a received user request, each of the at least one request concepts having an associated request concept score;

5) determining a similarity score between the ad and the received user request using, at least, the determined at least one targeting concept of the ad and its adjusted targeting concept score in combination with the at least one request concept and its request concept score; and

6) controlling serving of the ad using at least the determined similarity score.