Identifying negative keywords associated with advertisements

• US 8,229,942 B1
• Filed: 04/17/2007
• Issued: 07/24/2012
• Est. Priority Date: 04/17/2007
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Hubinette, Carl F.
• Primary Examiner(s): Coby, Frantz
• Application Number:
  US11/736,452
• Time in Patent Office: 1,925 Days
• Field of Search: 707/713, 707/721, 707/722, 707/736, 707/759, 707/769, 707/771, 707/773
• US Class Current: 707/759
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

The present disclosure includes a system and method for identifying negative keywords associated with advertisements. In some implementations, a method for advertising includes identifying one or more keywords associated with an advertisement. The keywords used to embed the advertisement in Web pages such as those including search results. By scoring keywords to identify those keywords that are not relevant to the web page content, one or more negative keywords can be identified in order to limit the embedding of non-relevant advertisements in web page content.

Claims

1. A computer-implemented method, comprising:

identifying a content item, the content item being associated with one or more keywords that are used in determining whether to embed the content item in web pages including search results;

extracting from a search log a plurality of previously-submitted search criteria that have resulted in presentation of the content item in respective search result web pages for the search criteria due to respective associations between the search criteria and the keywords of the content item;

identifying, from among the plurality of previously-submitted search criteria, at least one search criterion that is irrelevant to a topic of the content item, based on one or more evaluation criteria; and

evaluating, by one or more processors, respective recurrences of terms in the at least one irrelevant search criterion to identify a partial subset of the terms as one or more negative keywords for the content item,

wherein identifying the at least one irrelevant search criterion is further based on at least on metrics representing probabilities that terms of each of the plurality of previously-submitted search criteria occurs within one or more keyword clusters of the content item.

2. The method of claim 1, wherein identifying the at least one irrelevant search criterion is based at least in part on proximity of each of the plurality of previously-submitted search criteria to the topic of the content item.

3. The method of claim 1, wherein identifying the at least one irrelevant search criterion is based at least on user input representing user actions performed with respect to the content item.

4. The method of claim 1, further comprising:

determining a respective metric associated with each of one or more subsets of terms of each of the plurality of previously-submitted search criteria based, at least in part, on probabilities that the subset of terms occurs within the keyword clusters associated with the content item;

aggregating the respective metrics associated with the subsets of terms of each of the previously-submitted search criteria to obtain a respective aggregated metric for the previously-submitted search criterion; and

determining, for each of the plurality of previously-submitted search criteria, a respective proximity to the topic associated with the content item based, at least in part, on the respective aggregated metric of the previously-submitted search criterion.

5. The method of claim 4, further comprising:

identifying at least one proximity threshold associated with the topic of the content item; and

wherein identifying the at least one irrelevant search criterion is in response to at least the respective proximity of the at least one irrelevant search criterion violating the at least one proximity threshold.

6. The method of claim 1, wherein evaluating respective recurrences of terms in the at least irrelevant search criterion to identify the partial subset of the terms as one or more negative keywords for the content item further comprises:

identifying one or more terms that reoccur in the at least one irrelevant search criterion but are absent in other search criteria in the plurality of previously-submitted search criteria that have not been determined to be irrelevant to the topic of the content item; and

selecting the partial subset of the terms from the identified one or more recurring terms.

7. The method of claim 6, wherein selecting the partial subset of the terms from the identified one or more recurring terms further comprises:

for each of the one or more recurring terms;

aggregating respective metrics associated with one or more of the at least one irrelevant search criterion that include the recurring term to obtain an aggregated metric;

comparing the aggregated metrics with a threshold value; and

identifying the recurring term as one of the one or more negative keywords for the content item in response to the aggregated metric violating the threshold value.

8. The method of claim 7, wherein the respective metric associated with each of the at least one irrelevant search criterion is based at least on a measure of non-relevance of the search criterion relative to a keyword of the content item and an amount of impressions associated with the keyword of the content item.

9. A non-transitory computer-readable medium having instructions stored thereon, which, when executed by at least one processor, cause the processor to perform operations comprising:

identifying a content item, the content item being associated with one or more keywords that are used in determining whether to embed the content item in web pages including search results;

extracting from a search log a plurality of previously-submitted search criteria that have resulted in presentation of the content item in respective search result web pages for the search criteria due to respective associations between the search criteria and the keywords of the content item;

identifying, from among the plurality of previously-submitted search criteria, at least one search criterion that is irrelevant to a topic of the content item, based on one or more evaluation criteria; and

evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify a partial subset of the terms as one or more negative keywords for the content item,

wherein identifying the at least one irrelevant search criterion is further based at least on metrics representing probabilities that terms of each of the plurality of previously-submitted search criteria occur within one or more keyword clusters of the content item.

10. The computer-readable medium of claim 9 wherein identifying the at least one irrelevant search criterion is based at least in part on proximity of each of the plurality of previously-submitted search criteria to the topic of the content item.

11. The computer-readable medium of claim 9, wherein identifying the at least one irrelevant search criterion is based at least on user input representing user actions performed with respect to the content item.

12. The computer-readable medium of claim 9, wherein the operations further comprise:

determining a respective metric associated with each of one or more subsets of terms of each of the plurality of previously-submitted search criteria based, at least in part, on probabilities that the subset of terms occurs within the keyword clusters associated with the content item;

aggregating the respective metrics associated with the subsets of terms of each of the plurality of previously-submitted search criteria to obtain a respective aggregated metric for the previously-submitted search criterion; and

determining, for each of the plurality of previously-submitted search criteria, a respective proximity to the topic associated with the content item based, at least in part, on the aggregated metrics of the previously-submitted search criterion.

13. The computer-readable medium of claim 12, wherein the operations further comprise:

identifying at least one proximity threshold associated with the topic of the content item; and

wherein identifying the at least one irrelevant search criterion is in response to at least the respective proximity of the at least one irrelevant search criterion violating the at least one proximity threshold.

14. The computer-readable medium of claim 9, wherein evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify the partial subset of the terms as one or more negative keywords for the content item further comprises:

identifying one or more terms that reoccur in the at least one irrelevant search criterion but are absent in other search criteria in the plurality of previously-submitted search criteria that have not been determined to be to the topic of the content item; and

selecting the partial subset of the terms from the identified one or more recurring terms.

15. The computer-readable medium of claim 14, wherein selecting the partial subset of the terms from the identified one or more recurring terms further comprises:

for each of the one or more recurring terms;

aggregating respective metrics associated with one or more of the at least one irrelevant search criterion that include the recurring term;

comparing the aggregated metrics with a threshold value; and

identifying the recurring term as one of the one or more negative keywords for the content item in response to the aggregated metric violating the threshold value.

16. The computer-readable medium of claim 15, wherein the respective metric associated with each of the at least one irrelevant search criterion is based at least on a measure of non-relevance of the search criterion relative to a keyword of the content item and an amount of impressions associated with the keyword of the content item.

17. A server for identifying negative keywords, comprising one or more processors, a memory coupled to the one or more processors, and a network interface, comprising:

means for identifying a content item, the content item being associated with one or more keywords that are used in determining whether to embed the content item in web pages including search results;

means for extracting from a search log a plurality of previously-submitted search criteria that have resulted in presentation of the content item in respective search result web pages for the search criteria due to respective associations between the search criteria and the keywords of the content item;

means for identifying, from among the plurality of previously-submitted search criteria, at least one search criterion that is irrelevant to a topic of the content item, based on one or more evaluation criteria; and

means for evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify a partial subset of the terms as one or more negative keywords for the content item,

wherein identifying the at least one irrelevant search criterion is further based on at least metrics representing probabilities that terms of each of the plurality of previously-submitted search criteria occur within one or more keyword clusters of the content item.

18. The server of claim 17, wherein identifying the at least one irrelevant search criterion is based at least in part on proximity of each of the plurality of previously-submitted search criteria to the topic of the content item.

19. The server of claim 17, wherein identifying the at least one irrelevant search criterion is based at least on user input representing user actions performed with respect to the content item.

20. The server of claim 17, further comprising:

means for determining a respective metric associated with each of one or more subsets of terms of each of the plurality of previously-submitted search criteria based, at least in part, on probabilities that the subset of terms occurs within the keyword clusters associated with the content item;

means for aggregating the respective metrics associated with the subsets of terms of each of the plurality of previously-submitted search criteria to obtain a respective aggregated metric for the previously-submitted search criterion; and

means for determining, for each of the plurality of previously-submitted search criteria, a respective proximity to the topic associated with the content item based, at least in part, on the aggregated metrics of the previously-submitted search criterion.

21. The server of claim 20, further comprising:

means for identifying at least one proximity threshold associated with the topic of the content item; and

wherein identifying the at least one irrelevant search criterion is in response to at least the respective proximity of the at least one irrelevant search criterion violating the at least one proximity threshold.

22. The server of claim 17, wherein the means for evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify the partial subset of the terms as one or more negative keywords for the content item further comprises:

means for identifying one or more terms that reoccur in the at least one irrelevant search criterion but are absent in other search criteria in the plurality of previously-submitted search criteria that have not been determined to be irrelevant to the topic of the content item; and

means for selecting the partial subset of the terms from the identified one or more recurring terms.

23. The server of claim 22, wherein the means for selecting the partial subset of the terms from the identified one or more recurring terms further comprises:

for each of the one or more recurring terms;

means for aggregating respective metrics associated with one or more of the at least one irrelevant search criterion that include the recurring term to obtain an aggregated metric;

means for comparing the aggregated metrics with a threshold value; and

means for identifying the recurring term as one of the one or more negative keywords for the content item in response to the aggregated metric violating the threshold value.

24. The server of claim 23, wherein the respective metric associated with each of the at least one irrelevant search criterion is based at least on a measure of non-relevance of the search criterion relative to a keyword of the content item and an amount of impressions associated with the keyword of the content item.

25. A system, comprising:

one or more processors; and

memory having instructions stored thereon, the instructions, when executed by the one or more processors, cause the one or more processors to perform operations comprising;

identifying a content item, the content item being associated with one or more keywords that are used in determining whether to embed the content item in web pages including search results;

extracting from a search log a plurality of previously-submitted search criteria that have resulted in presentation of the content item in respective search result web pages for the search criteria due to respective associations between the search criteria and the keywords of the content item;

identifying, from among the plurality of previously-submitted search criteria, at least one search criterion that is irrelevant to a topic of the content item, based on one or more evaluation criteria; and

evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify a partial subset of the terms as one or more negative keywords for the content item,

wherein identifying the at least one irrelevant search criterion is further based at least on metrics representing probabilities that terms of each of the plurality of previously-submitted search criteria occur within one or more keyword clusters of the content item.

26. The system of claim 25, wherein identifying the at least one irrelevant search criterion is based at least in part on proximity of each of the plurality of previously submitted search criteria to the topic of the content item.

27. The system of claim 25, wherein identifying the at least one irrelevant search criterion is based at least on user input representing user actions performed with respect to the content item.

28. The system of claim 25, wherein the operations further comprise:

determining a respective metric associated with each of one or more subsets of terms of each of the plurality of previously-submitted search criteria based, at least in part, on probabilities that the subset of terms occurs within the keyword clusters associated with the content item;

aggregating the respective metrics associated with the subsets of terms of each of the previously-submitted search criteria to obtain a respective aggregated metric for the previously-submitted search criterion; and

determining, for each of the plurality of previously-submitted search criteria, a respective proximity to the topic associated with the content item based, at least in part, on the respective aggregated metric of the previously-submitted search criterion.

29. The system of claim 28, wherein the operations further comprise:

identifying at least one proximity threshold associated with the topic of the content item; and

wherein identifying the at least one irrelevant search criterion is in response to at least the respective proximity of the at least one irrelevant search criterion violating the at least one proximity threshold.

30. The server of claim 25, wherein evaluating respective recurrences of terms in the at least one irrelevant search criterion to identify the partial subset of the terms as one or more negative keywords for the content item further comprises:

identifying one or more terms that reoccur in the at least one irrelevant search criterion but are absent in other search criteria in the plurality of previously-submitted search criteria that have not been determined to be irrelevant to the topic of the content item; and

selecting the partial subset of the terms from the identified one or more recurring terms.

31. The system of claim 30, wherein selecting the partial subset of the terms from the identified one or more recurring terms further comprises:

for each of the one or more recurring terms;

aggregating respective metrics associated with one or more of the at least one irrelevant search criterion that include the recurring term to obtain an aggregated metric;

comparing the aggregated metric with a threshold value; and

identifying the recurring term as one of the one or more negative keywords for the content item in response to the aggregated metric violating the threshold value.

32. The system of claim 30, wherein the respective metric associated with each of the at least one irrelevant search criterion is based at least on a measure of non-relevance of the search criterion relative to a keyword of the content item and an amount of impressions associated with the keyword of the content item.

33. A computer-implemented method, comprising:

identifying a content item, the content item being associated with one or more keywords that are used for targeting delivery of the content item;

extracting from a search log a plurality of previously-submitted search keywords that have been resolved to the content item;

identifying, from among the plurality of previously submitted search keywords, multiple search criteria that are irrelevant to a topic of the content item, based on metrics representing probabilities that terms of each of the plurality of previously-submitted search keywords occur within one or more keyword clusters of the content item; and

evaluating respective recurrences of terms in the multiple irrelevant search criteria to identify a partial subset of the terms as one or more negative keywords for the content item.

34. The method of claim 33, wherein identifying the multiple irrelevant search criteria is based at least in part on proximity of each of the plurality of previously-submitted search criteria to the topic of the content item.

35. The method of claim 33, wherein identifying the multiple irrelevant search criteria is based at least on user input representing user actions performed with respect to the content item.

36. The method of claim 33, further comprising:

determining a respective metric associated with each of one or more subsets of terms of each of the plurality of previously-submitted search criteria based, at least in part, on probabilities that the subset of terms occurs within the keyword clusters associated with the content item;

aggregating the respective metrics associated with the subsets of terms of each of the plurality of previously-submitted search criteria to obtain a respective aggregated metric for the previously-submitted search criterion; and

determining, for each of the plurality of previously-submitted search criteria, a respective proximity to the topic associated with the content item based, at least in part, on the respective aggregated metric of the previously-submitted search criterion.

37. The method of claim 36, further comprising:

identifying at least one proximity threshold associated with the topic of the content item; and

wherein identifying the multiple irrelevant search criteria is in response to at least the respective proximities of the multiple irrelevant search criteria violating the at least one proximity threshold.

38. The method of claim 33, wherein evaluating respective recurrences of terms in the multiple irrelevant search criteria to identify the partial subset of the terms as one or more negative keywords for the content item further comprises:

identifying one or more terms that reoccur in the multiple irrelevant search criteria but are absent in other search criteria in the plurality of previously-submitted search criteria that have not been determined to be irrelevant to the topic of the content item; and

selecting the partial subset of the terms from the identified one or more recurring terms.

39. The method of claim 38, wherein selecting the partial subset of the terms from the identified one or more recurring terms further comprises:

for each of the one or more recurring terms;

aggregating respective metrics associated with one or more of the multiple irrelevant search criteria that include the recurring term to obtain an aggregated metric;

comparing the aggregated metric with a threshold value; and

identifying the recurring term as one of the one or more negative keywords for the content item in response to the aggregated metric violating the threshold value.

40. The method of claim 38, wherein the respective metric associated with each of the multiple irrelevant search criteria is based at least on a measure of non-relevance of the search criterion relative to a keyword of the content item and an amount of impressions associated with the keyword of the content item.