System and method for determining a composite score for categorized search results

• US 7,814,085 B1
• Filed: 02/26/2004
• Issued: 10/12/2010
• Est. Priority Date: 02/26/2004
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Pfleger, Karl, Larson, Brian
• Primary Examiner(s): Breene; John E
• Assistant Examiner(s): Colan; Giovanna
• Application Number:
  US10/787,648
• Time in Patent Office: 2,420 Days
• Field of Search: 707/3, 707/6, 707/5, 707/7, 707/1, 707/708, 707/710, 707/748, 707/749, 707/750, 704/7, 705/14
• US Class Current: 707/708
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A system and method for scoring documents is described. One or more documents are identified responsive to a search criteria. A text match score indicating a quality of match of the identified documents is determined. A category match score is determined over categories. A document-categories score is determined indicating a quality of match between an identified document and a plurality of categories. A search criteria-categories score is determined indicating a quality of match between the search criteria and the categories. An overall score is determined based on the text match score and the category match score.

Claims

1. A system comprising:

one or more computers; and

a nonvolatile storage coupled to the one or more computers and having instructions stored thereon which, when executed by the one or more computers, cause the one or more computers to perform operations comprising;

receiving a search criteria;

selecting one or more categories associated with the search criteria;

selecting one or more documents responsive to the search criteria;

calculating a text match score for each document based on degree of match between the responsive document and the search criteria;

calculating a document-categories score for each of the one or more categories based on a degree of match between each document and each of the one or more categories; and

calculating a search criteria-categories score for each of the one or more categories based on a degree of match between the search criteria and each of the one or more categories, wherein the search criteria-categories score for a particular category indicates the degree of match between the search criteria and the category;

calculating a category match score for each document by combining the document-categories score of each of the one or more categories and the respective search criteria-categories score;

calculating an overall score for each document based on the text match score of each document and the respective category match score;

ranking the one or more documents according to the respective overall scores; and

providing the ranked one or more documents.

2. A system according to claim 1, wherein the search criteria is a search query comprising one or more query terms, and

wherein the text match score indicates a quality of match between the query terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the query terms and each of the one or more categories.

3. A system according to claim 1, wherein the search criteria includes a concept, the concept comprising one or more terms, and

wherein the text match score indicates a quality of match between the terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the terms and each of the one or more categories.

4. A system according to claim 1, further comprising:

defining the categories by aggregating a plurality of documents in the text corpus.

5. A system according to claim 1, further comprising:

mapping one or more of the documents to one or more of the one or more categories; and

associating one or more association strengths to matches between each document and the one or more categories.

6. A system according to claim 5, wherein the instructions further comprise:

calculating the document-categories score based on the degree of match for each document using the association strengths.

7. A system according to claim 1, further comprising:

associating words and word phrases with the one or more categories; and

assigning association strengths to the word and word phrase associations.

8. A system according to claim 7, wherein the instructions further comprise:

calculating the search criteria-categories score using a degree of match for the search criteria based on the association strengths of the word and word phrase associations.

9. A system according to claim 1, wherein the instructions further comprise:

calculating an aggregate category match score using the document-categories score and the search criteria-categories score.

10. A system according to claim 9, wherein the instructions further comprise:

evaluating a dot product of the document-categories score and the search criteria-categories score.

11. A system according to claim 1, wherein the blending function includes a weighted combination of the text match score and the category match score.

12. A system according to claim 11, wherein the composite score is determined in accordance with the equation:

S=a S_t+β

·

S_e

where S is the composite score, Se is the text match score, S_c is the category match score and α and

β

are weighting factors.

13. A system according to claim 11, wherein the composite score is determined in accordance with the equation:

S=α

S_t+β

S_c+γ

S_c+y·

S_tS_c

where S is the composite score, S_t is the text match score, S_c is the category match score, α

, β and

y are weighing factors.

14. A system according to claim 1, wherein the instructions further comprise:

specifying the categories as at least one of a list, a hierarchy and a searchable form of data structure.

15. A system according to claim 1, wherein the categories are applicable to shopping with one or more of the categories defining a category of goods or services.

16. A method for scoring documents using a server, comprising:

receiving a search criteria at a search engine;

identifying one or more categories associated with the search criteria;

identifying one or more documents responsive to the search criteria;

calculating a text match score for each document based on a degree of match between the respective document and the search criteria;

calculating a document-categories score for each of the one or more categories based on a degree of match between each document and each of the one or more categories;

calculating a search criteria-categories score for each of the one or more categories based on a degree of match between the search criteria and each of the one or more categories, wherein the search criteria-categories score for a particular category indicates the degree of match between the search criteria and the category;

calculating, using the server, a category match score for each document by combining the document-categories score of each of the one or more categories and the respective search criteria-categories score;

calculating an overall score for each document based on the text match score of each document and the respective category match score;

ranking the one or more documents according to the respective overall scores; and

providing the ranked one or more documents using a network interface.

17. A method according to claim 16, wherein the search criteria is a search query comprising one or more query terms, and

wherein the text match score indicates a quality of match between the query terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the query terms and each of the one or more categories.

18. A method according to claim 16, wherein the search criteria includes a concept, the concept comprising one or more terms, and

wherein the text match score indicates a quality of match between the terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the terms and each of the one or more categories.

19. A method according to claim 16, further comprising:

defining the categories by aggregating a plurality of documents in the text corpus.

20. A method according to claim 16, further comprising:

mapping one or more of the one or more documents to one or more of the one or more categories; and

assigning one or more association strengths to matches between each document and the one or more categories.

21. A method according to claim 20, wherein calculating the document-categories score further comprises:

calculating the document-categories score based on the degree of match for each document using the association strengths.

22. A method according to claim 16, further comprising:

associating words and word phrases with the one or more categories; and

assigning association strengths to the word and word phrase associations.

23. A method according to claim 22, wherein calculating the search criteria-categories score comprises:

calculating the search criteria-categories score based on a qualitative degree of match for the search criteria based on the association strengths of the word and word phrase associations.

24. A method according to claim 16, further comprising:

calculating an aggregate category match score using the document-categories score and the search criteria-categories score.

25. A method according to claim 24, further comprising:

evaluating a dot product of the document-categories score and the search criteria-categories score as the aggregate category match score.

26. A method according to claim 16, wherein the blending function includes a weighted combination of the text match score and the category match score.

27. A method according to claim 26, further comprising:

determining the composite score in accordance with the equation;

S=α

S_t+β

·

S_c

where S is the composite score, S_t is the text match score, S_c is the category match score and α and

β

are weighting factors.

28. A method according to claim 26, further comprising:

determining the composite score in accordance with the equation;

S=α

S_t+β

S_c+γ

S_c+yS_tS_c

where S is the composite score, S_t is the text match score, S_c is the category match score, α

, β and

γ

are weighing factors.

29. A method according to claim 16, further comprising:

specifying the categories as at least one of a list, a hierarchy and a searchable form of data structure.

30. A method according to claim 16, wherein the categories are applicable to shopping with one or more of the categories defining a category of goods or services.

31. A nonvolatile storage encoded with a computer program comprising instructions that, when executed, operate to cause a computer to perform operations comprising:

receiving a search criteria;

identifying one or more categories associated with the search criteria;

identifying one or more documents responsive to the search criteria;

calculating a text match score for each document based on a degree of match between the respective document and the search criteria;

calculating a document-categories score for each of the one or more categories based on a degree of match between each document and each of the one or more categories;

calculating a search criteria-categories score for each of the one or more categories based on a degree of match between the search criteria and each of the one or more categories, wherein the search criteria-categories score for a particular category indicates the degree of match between the search criteria and the category;

calculating a category match score for each document by combining the document-categories score of each of the one or more categories and the respective search criteria-categories score;

calculating an overall score for each document based on the text match score of each document and the respective category match score;

ranking the one or more documents according to the respective overall scores; and

providing the ranked one or more documents.

32. A storage according to claim 31, wherein the search criteria is a search query comprising one or more query terms, and

wherein the text match score indicates a quality of match between the query terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the query terms and each of the one or more categories.

33. A storage according to claim 31, wherein the search criteria includes a concept, the concept comprising one or more terms, and

wherein the text match score indicates a quality of match between the terms and each document, and

wherein the search criteria-categories score indicates a quality of match between the terms and each of the one or more categories.

34. A storage according to claim 31, further comprising:

defining the categories by aggregating a plurality of documents in the text corpus.

35. A storage according to claim 31, further comprising:

mapping one or more of the documents to one or more of the one or more categories; and

associating one or more association strengths to matches between each document and the one or more categories.

36. A storage according to claim 35, wherein the instructions further comprise:

calculating the document-categories score based on the degree of match for each document using the association strengths.

37. A storage according to claim 31, further comprising:

associating words and word phrases with the one or more categories; and

assigning association strengths to the word and word phrase associations.

38. A storage according to claim 37, wherein the instructions further comprise:

calculating the search criteria-categories score using a degree of match for the search criteria based on the association strengths of the word and word phrase associations.

39. A storage according to claim 31, wherein the instructions further comprise:

calculating an aggregate category match score using the document-categories score and the search criteria-categories score.

40. A storage according to claim 39, wherein the instructions further comprise:

evaluating a dot product of the document-categories score and the search criteria-categories score.

41. A storage according to claim 31, wherein the blending function includes a weighted combination of the text match score and the category match score.

42. A storage according to claim 41, wherein the composite score is determined in accordance with the equation:

S=a St+J·

Se

where S is the composite score, Se is the text match score, Sc is the category match score and I and J are weighting factors.

43. A storage according to claim 41, wherein the composite score is determined in accordance with the equation:

S=I St+J Sc+K Sc+y. St Sc

where S is the composite score, St is the text match score, Sc is the category match score, I, J and y are weighing factors.

44. A storage according to claim 31, wherein the instructions further comprise:

specifying the categories as at least one of a list, a hierarchy and a searchable form of data structure.

45. A storage according to claim 31, wherein the categories are applicable to shopping with one or more of the categories defining a category of goods or services.