Query boosting based on classification

• US 7,440,968 B1
• Filed: 11/30/2004
• Issued: 10/21/2008
• Est. Priority Date: 11/30/2004
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Kamvar, Sepandar D., Oztekin, Bilgehan Uygar, Haveliwala, Taher H., Jeh, Glen, Macduff, Tessa
• Primary Examiner(s): Le; Debbie M
• Application Number:
  US11/001,403
• Time in Patent Office: 1,421 Days
• Field of Search: 707/3, 707/102, 707/10
• US Class Current: 1/1
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99937 Sorting
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

In a method of generating a search boost vector, a first set of content location identifiers for first content are provided. A second set of content location identifiers for second content are identified. The second content is related to the first content in accordance with one or more first categories corresponding to the first content. The one or more first categories may be determined by applying a text classifier to the first content. The second content may also be identified based on one or more user profiles and/or one or more content site profiles. The search boost vector is generated using weights corresponding to the content location identifiers in the first set of content location identifiers and/or the second set of content location identifiers. The one or more search boost vectors may be used to boost one or more search results.

Claims

1. A method, comprising:

prior to receiving a search query,

providing to a boost vector module a first set of content location identifiers for first content corresponding to one or more first categories;

selecting a second set of content location identifiers for second content based on the one or more first categories, wherein the second content is predefined with respect to the one or more first categories;

generating a personalized search boost vector including a set of weights corresponding to content location identifiers in the second set of content location identifiers;

wherein the search boost vector identifies a plurality of content location identifiers in the second set and includes a respective weight for each of the identified content location identifiers; and

storing the search boost vector in a computer readable storage medium.

2. The method of claim 1, further including:

after receiving the search query;

generating search results;

selecting at least one search boost vector, including the stored search boost vector, from a set of stored search boost vectors;

boosting, so as to reorder the search results, one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

providing boosted search results.

3. The method of claim 1, further comprising:

performing text classification on the first content to determine the one or more first categories corresponding to the first content.

4. The method of claim 3, wherein the one or more first categories are determined by applying a text classifier to the first content.

5. The method of claim 3, wherein the selecting identifying of the second set of content location identifiers includes accessing stored information relating predefined categories to content location identifiers.

6. The method of claim 1, further comprising:

determining weights corresponding to content location identifiers in a set selected from the group consisting of the first set of content location identifiers, the second set of content location identifiers and the first and second set of content location identifiers.

7. The method of claim 1, wherein the second set of content location identifiers are selected from one or more existing sets of content location identifiers corresponding to one or more of the first categories.

8. The method of claim 1, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more user profiles.

9. The method of claim 1, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more content site profiles.

10. A method, comprising:

prior to receiving a search query, storing a set of search boost vectors, wherein a respective search boost vector identifies a plurality of content location identifiers and includes a respective weight for each of the identified content location identifiers;

after receiving the search query;

receiving the search query;

generating search results;

selecting one or more search boost vectors from the set of stored search boost vectors in accordance with one or more categories, wherein the one or more categories correspond to the search query or a source of the search query;

boosting, so as to reorder the search results, one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

providing boosted search results.

11. The method of claim 10, wherein the selecting the one or more search boost vectors is further in accordance with one or more user profiles.

12. The method of claim 10, wherein the each search boost vector of the stored set of search boost vectors includes the set of content location identifiers corresponding to the set of weights in the search boost vector.

13. A computer-program product, comprising:

a computer readable storage medium and a computer program mechanism embedded therein, the computer program mechanism comprising;

instructions for providing to a boost vector module, prior to receiving a search query, a first set of content location identifiers for first content corresponding to one or more first categories;

instructions for selecting a second set of content location identifiers for second content based on the one or more first categories, wherein the second content is predefined with respect to the one or more first categories;

instructions for generating a personalized search boost vector including a set of weights corresponding to content location identifiers in the second set of content location identifiers;

wherein the search boost vector identifies a plurality of content location identifiers in the second set and includes a respective weight for each of the identified content location identifiers; and

instructions for storing the search boost vector in predetermined computer readable storage media.

14. The computer-program product of claim 13, further including:

instructions for generating search results;

instructions for selecting at least one search boost vector, including the stored search boost vector, from a set of stored search boost vectors;

instructions for boosting, so as to reorder the search results, one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

instructions for providing boosted search results.

15. The computer-program product of claim 13, further comprising:

instructions for performing text classification on the first content to determine the one or more first categories corresponding to the first content.

16. The computer-program product of claim 15, wherein the one or more first categories are determined by applying a text classifier to the first content.

17. The computer-program product of claim 15, wherein instructions for the selecting of the second set of content location identifiers includes accessing stored information relating predefined categories to content location identifiers.

18. The computer-program product of claim 13, further comprising:

instructions for determining weights corresponding to content location identifiers in a set selected from the group consisting of the first set of content location identifiers, the second set of content location identifiers and the first and second set of content location identifiers.

19. The computer-program product of claim 13, wherein the second set of content location identifiers are selected from one or more existing sets of content location identifiers corresponding to one or more of the first categories.

20. The computer-program product of claim 13, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more user profiles.

21. The computer-program product of claim 13, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more content site profiles.

22. A computer-program product, comprising:

a computer readable storage medium and a computer program mechanism embedded therein, the computer program mechanism comprising;

instructions for storing a set of search boost vectors prior to receiving a search query, wherein a respective search boost vector identifies a plurality of content location identifiers and includes a respective weight for each of the identified content location identifiers;

instructions for receiving the search query;

instructions for generating search results after receiving the search query;

instructions for selecting one or more search boost vectors from the set of stored search boost vectors in accordance with one or more categories, wherein the one or more categories correspond to the search query or a source of the search query;

instructions for boosting one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

instructions for providing boosted search results.

23. The computer-program product of claim 22, wherein the selecting the one or more search boost vectors is further in accordance with one or more user profiles.

24. The computer-program product of claim 22, wherein the selecting the one or more search boost vectors is further in accordance with one or more content site profiles, wherein each content site profile comprises one or more of a term-based profile, a link-based profile, and a category-based profile.

25. A computer, comprising:

a main memory;

a processor; and

a program, stored in the main memory and executed by the processor, the program including instructions for providing to a boost vector module, prior to receiving a search query, a first set of content location identifiers for first content corresponding to one or more first categories;

instructions for selecting a second set of content location identifiers for second content based on one the one or more first categories, wherein the second content is predefined with respect to the one or more first categories;

instructions for generating a personalized search boost vector including a set of weights corresponding to content location identifiers in the second set of content location identifiers;

wherein the search boost vector identifies a plurality of content location identifiers in the second set and includes a respective weight for each of the identified content location identifiers; and

instructions for storing the search boost vector in a computer readable storage medium.

26. The computer of claim 25, further including:

instructions for generating search results;

instructions for selecting at least one search boost vector, including the stored search boost vector, from a set of stored search boost vectors;

instructions for boosting, so as to reorder the search results, one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

instructions for providing boosted search results.

27. The computer of claim 25, further comprising:

instructions for performing text classification on the first content to determine the one or more first categories corresponding to the first content.

28. The computer of claim 27, wherein the one or more first categories are determined by applying a text classifier to the first content.

29. The computer of claim 27, wherein instructions for the selecting of the second set of content location identifiers includes accessing stored information relating predefined categories to content location identifiers.

30. The computer of claim 25, further comprising:

instructions for determining weights corresponding to content location identifiers in a set selected from the group consisting of the first set of content location identifiers, the second set of content location identifiers and the first and second set of content location identifiers.

31. The computer of claim 25, wherein the second set of content location identifiers are selected from one or more existing sets of content location identifiers corresponding to one or more of the first categories.

32. The computer of claim 25, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more user profiles.

33. The computer of claim 25, wherein the selecting the second set of content location identifiers for the second content is further in accordance with one or more content site profiles.

34. A computer, comprising:

a main memory;

a processor; and

a program, stored in the main memory and executed by the processor, the program including instructions for storing a set of search boost vectors prior to receiving a search query, wherein a respective search boost vector identifies a plurality of content location identifiers and includes a respective weight for each of the identified content location identifiers;

instructions for receiving the search query;

instructions for generating search results after receiving the search query;

instructions for selecting one or more search boost vectors from the set of search boost vectors;

in accordance with one or more categories;

wherein the one or more categories correspond to the search query or a source of the search query;

instructions for boosting one or more of the search results using the one or more search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

instructions for providing boosted search results.

35. The computer of claim 34, wherein the selecting the one or more search boost vectors is further in accordance with one or more user profiles.

36. The computer of claim 34, wherein the selecting the one or more search boost vectors is further in accordance with one or more content site profiles, wherein each content site profile comprises one or more of a term-based profile, a link-based profile, and a category-based profile.

37. A computer, comprising:

a main memory;

a processor; and

a program, stored in the main memory and executed by the processor, the program including a boost vector module;

a means for providing to the boost vector module a first set of content location identifiers for first content corresponding to one or more first categories prior to receiving a search query;

a means for selecting a second set of content location identifiers for second content based on the one or more first categories, wherein the second content is predefined with respect to the one or more first categories; and

a means for generating a personalized search boost vector including a set of weights corresponding to content location identifiers in the second set of content location identifiers;

wherein the search boost vector identifies a plurality of content location identifiers in the second set and includes a respective weight for each of the identified content location identifiers; and

a means for storing the search boost vector in a computer readable storage medium.

38. A computer, comprising:

a main memory;

a processor; and

a program, stored in the main memory and executed by the processor, the program including a means for storing a set of search boost vectors prior to receiving a search query, wherein a respective search boost vector identifies a plurality of content location identifiers and includes a respective weight for each of the identified content location identifiers;

a means for generating search results after receiving the search query;

a means for selecting one or more search boost vectors from the set of stored search boost vectors in accordance with one or more categories, where in the one or more categories correspond to the search query or a source of the search query;

a means for boosting, so as to reorder the search results, one or more of the search results using the one or more selected search boost vectors;

wherein each boosted search result matches a respective content location identifier in at least one of the selected search boost vectors; and

a means for providing boosted search results.