Advertisement topic diversification and ranking

• US 8,468,083 B1
• Filed: 08/10/2007
• Issued: 06/18/2013
• Est. Priority Date: 08/10/2007
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Szulczewski, Piotr
• Primary Examiner(s): Kazimi, Hani M
• Assistant Examiner(s): Graham, Clement B
• Application Number:
  US11/837,070
• Time in Patent Office: 2,139 Days
• Field of Search: 705 35- 38
• US Class Current: 705/37
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

One or more first groups of one or more first content items are identified. A first score is determined for each of the one or more first groups. A second group of one or more second content items is identified. At least one of the one or more second groups correspond to at least one of the one or more first groups. A second score is determined for each of the one or more second groups. A third score is determined for the first groups and the corresponding second group based on the first scores and second scores. An adjustment is made to a ranking of the one or more first content items based on the third scores.

Claims

1. A method implemented by a computer, the method comprising:

identifying, by a computer, first groups of one or more first content items, wherein each of the first groups is defined by one or more first content items grouped under a context corresponding to the first group, and each first group has a different corresponding context;

generating, by the computer, a first score for each of the one or more first groups;

identifying, by the computer, second groups of one or more second content items, wherein each of the second groups is defined by one or more second content items grouped under a context corresponding to the second group, and each second group has a different corresponding context;

generating by the computer, a second score for each of the one or more second groups;

identifying, by the computer, corresponding group pairs, each corresponding group pair being a first group and a second group having a same corresponding context;

generating, by the computer, third scores for the corresponding group pairs, each third score being for a corresponding group pair and being based on the respective first score and second score of the first group and the second group of the corresponding group pair;

adjusting, by the computer, a ranking of the one or more first content items based on the third scores to create a diversified ranking in which a first content item grouped under a first context corresponding to a first group and a first content item grouped under a second context corresponding to a second group are presented with one or more second content items;

providing data, by the computer, of the diversified ranking of the one or more first content items.

2. The method of claim 1, further comprising:

identifying the one or more first content items based on a web page content.

3. The method of claim 1, further comprising:

identifying the one or more second content items based on a web page content.

4. The method of claim 1, further comprising:

identifying the one or more first content items based on a search query.

5. The method of claim 1, further comprising:

identifying the one or more second content items based on a search query.

6. The method of claim 1, wherein generating a first score comprises:

assigning first content scores to each of the first content items that is substantially inversely proportional to a rank of the first content item;

adding the first content scores of same grouped first content items together.

7. The method of claim 1, wherein generating a second score comprises:

assigning second content scores to each of the second content items that is substantially inversely proportional to a rank of the second content item; and

adding the second content scores of same grouped second content itemstogether.

8. The method of claim 1, wherein each of the first content items is an advertisement.

9. The method of claim 1, wherein each of the second content items is a search result.

10. A method implemented by a computer, the method comprising:

identifying, by a computer, first groups of one or more first content items, each first group defined by one or more first content items that are grouped under a context corresponding to the first group, and wherein each first group has a different corresponding context;

identifying, by the computer, second groups, each second group defined by one or more second content items that are grouped under a context corresponding to the second group, and wherein each second group has a different corresponding context;

scoring, by the computer, the one or more first content items in each of the first groups and the one or more second content items in each of the second groups;

scoring, by the computer, the first groups and the second groups;

identifying, by the computer, corresponding group pairs, each corresponding group pair being a first group and a second group having the same corresponding context;

generating, by the computer, combined group scores, each combined group score being generated for each corresponding group pair, the combined group score being based on the respective first score and second score of the first group and the second group of the corresponding group pair;

determining, by the computer, whether to diversify a ranking of the one or more first content items based on the combined group scores wherein a first content item grouped under a first context corresponding to a first group and a first content item grouped under a second context corresponding to a second group are presented with one or more second content items;

in response to determining to diversify the ranking of the one or more first content items, providing, by a computer, data of a diversified ranking of the one or more first content items.

11. The method of claim 10, wherein at least one of the one or more second groups correspond to at least one of the one or more first groups.

12. The method of claim 10, wherein scoring, by the computer, the first groups and the second groups comprises:

assigning to each of the first groups a value derived from the sum of the one or more first content items that are grouped under a context corresponding to respective first group;

assigning to each of the second groups a value derived from the sum of the one or more second content items that are grouped under a context corresponding to respective second group.

13. The method of claim 10, wherein scoring, by the computer, the first groups and the second groups comprises:

determining first scores for each of the one or more first content items in each of the first groups, each first score substantially inversely proportional to a ranking of the one or more first content items;

determining second scores for each of the one or more second content items in each of the second groups, each second score substantially inversely proportional to a ranking of the one or more second content items.

14. The method of claim 13, wherein determining first scores for each of the one or more first content items comprises:

assigning first content scores to each of the first content items that is substantially inversely proportional to a rank of the first content item; and

adding the first content scores together.

15. The method of claim 13, wherein determining second scores for each of the one or more second content items comprises:

assigning second content scores to each of the second content items that is substantially inversely proportional to a rank of the second content item; and

adding the second content scores together.

16. A method implemented by a computer, the method comprising:

receiving, by a computer, advertisements responsive to a search query;

receiving, by the computer, search results responsive to the search query;

identifying, by the computer, advertisement context clusters for the advertisements, each advertisement context cluster defined by one or more of the advertisements that are grouped under a topic corresponding to the advertisement context cluster, and wherein each advertisement context cluster has a different corresponding topic;

identifying, by the computer, search result context clusters for the search results, each search result context cluster defined by one or more of the search results that are grouped under a topic corresponding to the search result context cluster, and wherein each search result cluster has a different corresponding topic;

determining, by the computer, a score for each of the advertisement context clusters and the search result context clusters;

identifying, by the computer, corresponding cluster pairs, each corresponding cluster pair being an advertisement context cluster and a search result context cluster having the same corresponding topic;

generating, by the computer, combined cluster scores for the corresponding cluster pairs, each combined cluster score being for a corresponding cluster pair and being based on a respective score for an advertisement context cluster of the corresponding cluster pair and a respective score for a search result context cluster of the corresponding cluster pair;

adjusting, by the computer, a presentation ranking of the advertisements based on the combined cluster scores, wherein the presentation ranking of the advertisements is diversified according to the combined cluster scores so that at least a first advertisement grouped under a first topic and at least a second advertisement grouped under a second topic are presented with the search results;

providing data, by the computer, that cause presentation of diversified advertisements on a client device.

17. A method implemented by a computer, the method comprising:

receiving, by a computer, one or more advertisements and one or more search results, wherein the one or more advertisements and one or more search results are each associated with a rank;

mapping, by the computer, the one or more advertisements to one or more advertisement clusters, wherein each advertisement cluster has a different context;

assigning, by the computer, an advertisement score to each of the advertisements that is substantially inversely proportional to a rank of the advertisements; and

determining, by the computer, advertisement cluster scores by adding advertisement scores of the advertisements in each of the advertisement clusters together;

mapping, by the computer, the one or more search results to one or more search clusters, wherein each search cluster has a different context;

assigning, by the computer, a search score to each of the search results that is substantially inversely proportional to a rank of the search result;

determining, by the computer, search cluster scores by adding search scores of the search results in each of the search clusters together;

identifying, by the computer, corresponding group pairs, each corresponding group pair being an advertisement cluster and a search cluster having the same corresponding context;

generating, by the computer, a context score by combining an advertisement cluster score for an advertisement cluster of a corresponding group pair and a search cluster score for a search cluster of the corresponding group pair;

producing, by the computer, a diversified ranking of the one or more advertisements based on the context score wherein a first advertisement clustered under a first context and a second advertisement grouped under a second context are presented with the one or more search results;

providing data, by the computer, of the diversified ranking of the one or more advertisements.

18. The method of claim 17, wherein the one or more advertisements are returned in response to a search query.

19. The method of claim 17, wherein the one or more search results are returned in response to a search query.

20. A system, comprising:

a clustering engine, implemented as part of a system comprising one or more computing devices, that identifies one or more first groups of one or more same context first content items, with each of the one or more first groups having a different corresponding context, and identifies one or more second groups of one or more same context second content items, with each of the one or more second content items having a different corresponding context, and identifies at least one or more corresponding group pairs, each corresponding group pair being a first group and a second group having the same corresponding context; and

a scoring engine, implemented as part of the system comprising one or more computing devices, that determines a first score for each of the one or more first groups, determines a second score for each of the one or more second groups, determines third scores for each corresponding group pair based on the respective first scores and second scores of the corresponding group pair, and provides diversified ranking of the one or more first content items based on the third scores.

21. The system of claim 20, wherein the clustering engine:

identifies the one or more first content items based on a web page content.

22. The system of claim 20, wherein the clustering engine:

identifies the one or more second content items based on a web page content.

23. The system of claim 20, wherein the clustering engine:

identifies the one or more first content items based on a search query.

24. The system of claim 20, wherein the clustering engine:

identifies the one or more second content items based on a search query.

25. A system, comprising:

a data processing apparatus configured to perform operations comprising;

identifying one or more first groups of one or more first content items, wherein each first group has a different corresponding topic;

identifying one or more second groups of one or more second content items, wherein the one or more second groups correspond to the one or more first groups, and wherein each second group has a different corresponding topic;

ranking the one or more first groups and the one or more second groups;

identifying, by the computer, corresponding group pairs, each corresponding group pair being a first group and a second group having a same corresponding topic;

generating a combined group score for each corresponding group pair, each combined group score based on the respective first score and second score of the first group and the second group of the corresponding group pair; and

diversifying the one or more first content items based on the combined group score.