System and method for providing text summarization for use in web-based content

• US 7,587,309 B1
• Filed: 12/01/2003
• Issued: 09/08/2009
• Est. Priority Date: 12/01/2003
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Rohrs, Christopher, Brants, Thorsten
• Primary Examiner(s): Hudspeth; David R
• Assistant Examiner(s): Rider; Justin W
• Application Number:
  US10/725,883
• Time in Patent Office: 2,108 Days
• Field of Search: 704/10, 715/236, 715/244
• US Class Current: 704/10
• CPC Class Codes: G06F 16/345 Summarisation for human users
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

A system and method for providing text summarization for use in Web-based content is presented. Text is determined responsive to an executed query. Phrases within the text are identified, and words within the phrases are marked using matches of the words within the phrases with words of the executed query and/or a format rule. Marked words are placed into the summarized text subject to space restrictions. A system and method for building Web-based advertising creatives is also presented. At least one item description responsive to an executed query is identified and a name is extracted. Marked words are placed into the advertising creative subject to space restrictions.

Claims

1. A system for providing text summarization for use in Web-based content, comprising:

a text identifier to determine text responsive to an executed query comprising one or more query terms;

a phrase identifier to identify at least one phrase within the text; and

a phrase summarizer to summarize at least one of the at least one phrase into summarized text, the phrase summarizer including

a word marker to mark at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

a word placer to place the at least one marked word into the summarized text subject to space restrictions.

2. A system according to claim 1, wherein the at least one phrase is identified by content separators evaluated within the text.

3. A system according to claim 2, wherein the content separators comprises at least one of breaking punctuation, parenthetical text and breaking prepositions.

4. A system according to claim 2, further comprising:

a phrase matcher to match at least one word in the at least one of the at least one phrase with at least one word of the executed query independent of at least one of case and suffixes.

5. A system according to claim 2, further comprising:

a word locater to locate at least one significant word situated relative to one such matched word within the at least one phrase, and to place the at least one significant word into the summarized text subject to space restrictions.

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

a word marker to mark one or more unplaced words situated relative to one or more matched words and the at least one significant word within the at least one phrase, and to place one or more marked words or matched words into the summarized text subject to space restrictions.

7. A system according to claim 6, further comprising:

a word selector to select the marked words in one of right-to-left and left-to-right order.

8. A system according to claim 1, wherein the at least one phrase is identified by parts of speech determined by tagging the parts of speech within the text.

9. A system according to claim 8, further comprising:

a word dropper to drop one or more words situated relative to a grammatical phrase boundary.

10. A system according to claim 9, further comprising:

a word adder to add one or more words situated relative to one or more non-dropped words.

11. A system according to claim 9, further comprising:

a boundary identifier to determine the grammatical phrase boundary based on at least one of one or more predetermined punctuation marks and a prepositional phrase.

12. A system according to claim 11, wherein the predetermined punctuation marks comprise at least one of a hyphen, colon and parenthesis.

13. A system according to claim 8, further comprising:

a word selector to select one or more words situated relative to a grammatical phrase boundary, wherein the word comprises one of a rightmost proper noun, a regular noun immediately following the rightmost proper noun, and a regular noun if the rightmost proper noun is not found, and to place the one or more selected words into the summarized text subject to space restrictions.

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

a confidence level associated with summarized text; and

an evaluator to evaluate a quality of the summarized text based on the confidence level.

15. A system according to claim 14, further comprising:

an adjuster to adjust the confidence level relative to one or more dropped words.

16. A system according to claim 1, wherein at least one word comprises a compound word comprising a plurality of individual words.

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

a case normalizer to normalize the case for letters occurring in each of the zero or more phrases.

18. A system according to claim 1, wherein the text comprises at least one of an advertisement, search results and Web content.

19. A system according to claim 1, wherein the text comprises non-Web content.

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

a text placer to place the summarized text in at least one of an advertising creative, summarized search results and summarized Web content.

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

a text placer to place the summarized text into at least one of a table cell and column entry.

22. A computer-implemented method for providing text summarization for use in Web-based content, comprising:

determining, using a computer system including one or more computing devices, text responsive to an executed query comprising one or more query terms and identifying at least one phrase within the text;

summarizing, using the computer system, at least one of the at least one phrase into summarized text, the act of summarizing including

marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

placing the at least one marked word into the summarized text subject to space restrictions; and

outputting at least some of the summarized text to induce rendering on a networked client device.

23. A computer-implemented method according to claim 22, further comprising:

identifying, using the computer system, the at least one phrase by content separators evaluated within the text.

24. A computer-implemented method according to claim 23, wherein the content separators comprises at least one of breaking punctuation, parenthetical text and breaking prepositions.

25. A computer-implemented method according to claim 23, further comprising:

matching, using the computer system, at least one word in the at least one of the at least one phrase with at least one word of the executed query independent of at least one of case and suffixes.

26. A computer-implemented method according to claim 23, further comprising:

locating, using the computer system, at least one significant word situated relative to one such matched word within the at least one phrase; and

placing, using the computer system, the at least one significant word into the summarized text subject to space restrictions.

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

marking, using the computer system, one or more unplaced words situated relative to one or more matched words and the at least one significant word within the at least one phrase; and

placing one or more marked words or matched words into the summarized text subject to space restrictions.

28. A computer-imlemented method according to claim 27, further comprising:

selecting, using the computer system, the marked words in one of right-to-left and left-to-right order.

29. A computer-imlemented method according to claim 22, further comprising:

identifying, using the computer system, the at least one phrase by parts of speech determined by tagging the parts of speech within the text.

30. A computer-imlemented method according to claim 29, further comprising:

dropping, using the computer system, one or more words situated relative to a grammatical phrase boundary.

31. A computer-imlemented method according to claim 30, further comprising:

adding, using the computer system, one or more words situated relative to one or more non-dropped words.

32. A computer-imlemented method according to claim 30, further comprising:

determining, using the computer system, the grammatical phrase boundary based on at least one of one or more predetermined punctuation marks and a prepositional phrase.

33. A computer-imlemented method according to claim 32, wherein the predetermined punctuation marks comprise at least one of a hyphen, colon and parenthesis.

34. A computer-imlemented method according to claim 30, further comprising:

selecting, using the computer system, one or more words situated relative to a grammatical phrase boundary, wherein the word comprises one of a rightmost proper noun, a regular noun immediately following the rightmost proper noun, and a regular noun if the rightmost proper noun is not found; and

placing, using the computer system, the one or more selected words into the summarized text subject to space restrictions.

35. A computer-imlemented method according to claim 22, further comprising:

associating, using the computer system, a confidence level with summarized text; and

evaluating, using the computer system, a quality of the summarized text based on the confidence level.

36. A computer-imlemented method according to claim 35, further comprising:

adjusting, using the computer system, the confidence level relative to one or more dropped words.

37. A computer-imlemented method according to claim 22, wherein at least one word comprises a compound word comprising a plurality of individual words.

38. A computer-imlemented method according to claim 22, further comprising:

normalizing, using the computer system, the case for letters occurring in each of the zero or more phrases.

39. A computer-imlemented method according to claim 22, wherein the text comprises at least one of an advertisement, search results and Web content.

40. A computer-imlemented method according to claim 22, wherein the text comprises non-Web content.

41. A computer-imlemented method according to claim 22, further comprising:

placing, using the computer system, the summarized text in at least one of an advertising creative, summarized search results and summarized Web content.

42. A computer-imlemented method according to claim 22, further comprising:

placing, using the computer system, the summarized text into at least one of a table cell and column entry.

43. A computer-imlemented method according to claim 22, wherein the computer system includes an ad server, the computer-implemented method further comprising:

providing, using the computer system, the summarized text to the ad server as an advertising creative; and

serving, using the computer system, the advertising creative to induce rendering of the advertising creative on a client device.

44. A computer-readable storage medium holding code for performing a method comprising:

determining, using a computer system including one or more computing devices, text responsive to an executed query comprising one or more query terms and identifying at least one phrase within the text;

summarizing, using the computer system, at least one of the at least one phrase into summarized text, the act of summarizing including

marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

placing the at least one marked word into the summarized text subject to space restrictions; and

outputting at least some of the summarized text to induce rendering on a networked client device.

45. An apparatus for providing text summarization for use in Web-based content, comprising:

means for determining text responsive to an executed query comprising one or more query terms and means for identifying at least one phrase within the text; and

means for summarizing at least one of the at least one phrase into summarized text, the means for summarizing including

means for marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

means for placing the at least one marked word into the summarized text subject to space restrictions.

46. A system for building Web-based advertising creatives, comprising:

an advertising server to identify at least one item description responsive to an executed query comprising one or more query terms;

a phrase identifier to extract a name from the item description and to identify at least one phrase within the name; and

a phrase summarizer to summarize at least one of the at least one phrase into an advertising creative, the phrase summarizer including

a word marker to mark at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

a word placer to place the at least one marked word into the advertising creative subject to space restrictions.

47. A system according to claim 46, further comprising:

a description identifier to retrieve a description from the item description and to supplement the advertising creative with the description subject to space restrictions.

48. A system according to claim 47, further comprising:

a description summarizer to summarize the description.

49. A system according to claim 46, further comprising:

a parse tree formed from the item description;

a parser to traverse the parse tree for constituents meeting syntactic constraints; and

a description identifier to supplement the advertising creative with the constituents subject to space restrictions.

50. A system according to claim 46, further comprising:

a category identifier to retrieve a category name associated with the item description and supplementing the advertising creative with the category name subject to space restrictions.

51. A system according to claim 50, further comprising:

a hierarchy of categories; and

a category summarizer to summarize the category name by prepending ancestor categories determined from the hierarchy.

52. A system according to claim 51, wherein redundant categories are removed from the category name.

53. A system according to claim 46, further comprising:

a confidence level associated with the category name; and

a category summarizer to select the category name having a substantially best confidence level.

54. A system according to claim 46, further comprising:

a merchant identifier to retrieve a merchant information from the item description and supplementing the advertising creative with the merchant information subject to space restrictions.

55. A system according to claim 46, further comprising:

a case normalizer to normalize the case for letters occurring in each of the zero or more words.

56. A system according to claim 46, wherein the item description comprises at least one of structured and unstructured Web content.

57. A system according to claim 46, wherein the item description comprises at least one of a product description and a service description.

58. A computer-imlemented method for building Web-based advertising creatives, comprising:

identifying, using a computer system including one or more computing devices, at least one item description responsive to an executed query comprising one or more query terms;

extracting, using the computer system, a name from the item description;

identifying, using the computer system, at least one phrase within the name;

summarizing, using the computer system, at least one of the at least one phrase into an advertising creative, the act of summarizing including

marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

placing the at least one marked word into the advertising creative subject to space restrictions; and

outputting the advertising creative to induce rendering on a networked client device.

59. A computer-imlemented method according to claim 58, further comprising:

retrieving, using the computer system, a description from the item description; and

supplementing, using the computer system, the advertising creative with the description subject to space restrictions.

60. A computer-imlemented method according to claim 59, further comprising:

summarizing, using the computer system, the description.

61. A computer-imlemented method according to claim 58, further comprising:

forming, using the computer system, the item description into a parse tree;

traversing, using the computer system, the parse tree for constituents meeting syntactic constraints; and

supplementing, using the computer system, the advertising creative with the constituents subject to space restrictions.

62. A computer-imlemented method according to claim 58, further comprising:

retrieving, using the computer system, a category name associated with the item description; and

supplementing, using the computer system, the advertising creative with the category name subject to space restrictions.

63. A computer-imlemented method according to claim 62, further comprising:

defining, using the computer system, a hierarchy of categories; and

summarizing, using the computer system, the category name by prepending ancestor categories determined from the hierarchy.

64. A computer-imlemented method according to claim 63, further comprising:

removing, using the computer system, redundant categories from the category name.

65. A computer-imlemented method according to claim 58, further comprising:

associating, using the computer system, a confidence level with the category name; and

selecting, using the computer system, the category name having a substantially best confidence level.

66. A computer-imlemented method according to claim 58, further comprising:

retrieving, using the computer system, a merchant information from the item description; and

supplementing, using the computer system, the advertising creative with the merchant information subject to space restrictions.

67. A computer-imlemented method according to claim 58, further comprising:

normalizing, using the computer system, the case for letters occurring in each of the zero or more words.

68. A computer-imlemented method according to claim 58, wherein the item description comprises at least one of structured and unstructured Web content.

69. A computer-imlemented method according to claim 58, wherein the computer system includes an ad server, the computer-implemented method further comprising:

providing, using the computer system, the advertising creative to the ad server; and

serving, using the computer system, the advertising creative to induce rendering of the advertising creative on a client device.

70. A computer-imlemented method according to claim 58, wherein the item description comprises at least one of a product description and a service description.

71. A computer-readable storage medium holding code for performing a method comprising:

identifying, using a computer system including one or more computing devices, at least one item description responsive to an executed query comprising one or more query terms;

extracting, using the computer system, a name from the item description;

identifying, using the computer system, at least one phrase within the name;

summarizing, using the computer system, at least one of the at least one phrase into an advertising creative, the act of summarizing including

marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

placing the at least one marked word into the advertising creative subject to space restrictions; and

outputting the advertising creative to induce rendering on a networked client device.

72. An apparatus for building Web-based advertising creatives, comprising:

means for identifying at least one item description responsive to an executed query comprising one or more query terms;

means for extracting a name from the item description;

means for identifying at least one phrase within the name; and

means for summarizing at least one of the at least one phrase into an advertising creative, the means for summarizing including

means for marking at least one word in the at least one of the at least one phrase using at least one of (A) a match of the at least one word with at least one word of the executed query, and (B) a format rule, and

means for placing the at least one marked word into the advertising creative subject to space restrictions.