Contextual relevance engine and knowledge delivery system

• US 7,139,757 B1
• Filed: 12/21/2001
• Issued: 11/21/2006
• Est. Priority Date: 12/21/2001
• Status: Active Grant

• Current Assignee: The Procter & Gamble Company
• Original Assignee: The Procter & Gamble Company
• Inventors: Tenenbaum, Ray, Hoffert, Eric, Apollonsky, Alex, Braunig, Warren, Cremin, Karl
• Primary Examiner(s): Rones, Charles
• Assistant Examiner(s): ORTIZ DITREN, BELIX M
• Application Number:
  US10/027,471
• Time in Patent Office: 1,796 Days
• Field of Search: 707/7, 707/3, 707/4, 707/5, 707/10, 707/6, 707/100, 707/103, 707/200, 707/575, 707/104.1, 707/102, 707/104, 707/201, 707/202, 707/511
• US Class Current: 1/1
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: Y10S 707/99933 Query processing, i.e. searching
• CPC Class Codes: Y10S 707/99937 Sorting
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A method and apparatus are described for providing a user with contextually relevant content from a library of information. According to one embodiment of the present invention, metadata tagged information about a user'"'"'s work profile and working context, and content within a library of information are read. The metadata tagged information about content within the library of information to metadata tagged information about one or more of the user'"'"'s work profile, the user'"'"'s working context, or other content within the library of information is matched using one or more programmable rules. A rank ordered set of the most contextually relevant content from the library of information is then delivered to the user based on the matching of the metadata tagged information.

Claims

1. A method comprising the steps of:

maintaining a library of information;

reading metadata tagged information about a user'"'"'s work profile and working context and content within the library of information;

matching the metadata tagged information about content within the library of information to metadata tagged information about one or more of the user'"'"'s work profile, the user'"'"'s working context, or other content within the library of information using one or more programmable rules; and

delivering to the user a rank ordered set of the most contextually relevant content from the library of information based on the matching of the metadata tagged information, wherein the one or more programmable rules each have a unique identifier (rule ID), a name, and a set of clauses defining the operations of that rule.

2. The method of claim 1, wherein each clause of the set of clauses comprises two or more of:

a name identifying the metadata field to be matched or operator to be performed;

a weight to be applied to the metadata field if a match is found;

an operator to act on the metadata field to be matched; and

a value for a field representing context.

3. The method of claim 2, wherein a clause comprises all of a name, a weight, an operator, and a value.

4. The method of claim 2, wherein the clause comprises a value forming a portion of a Boolean expression, and a name representing an operator used to combine all of the clauses.

5. The method of claim 4, wherein the operator comprises a Boolean operator.

6. The method of claim 4, wherein the operator comprises a “

matches all”

match that compares two or more values and excludes a document associated with the value if non-matching values are found.

7. The method of claim 4, wherein the operator comprises a “

matches”

match that compares two or more values, accumulates a score for matches found, and excludes a document associated with the values if no matches are found.

8. The method of claim 4, wherein the operator comprises an “

optional”

match that compares two or more values and accumulates a score for matches found.

9. The method of claim 4, wherein the operator comprises an “

optional text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions and accumulates a score for matches found.

10. The method of claim 4, wherein the operator comprises a “

matches text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions, accumulates a score for matches found, and excludes a document associated with the values if no matches are found.

11. The method of claim 4, further comprising:

generating a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching;

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

12. The method of claim 11, further comprising sorting documents deemed to be relevant, by relevance, from most relevant to least relevant, and presenting, for display to the user, only a top tier of documents deemed to be most relevant.

13. The method of claim 12, wherein any document that matches itself in the list of relevant documents is discarded and is not presented for display to the user.

14. The method of claim 11, wherein each weight is defined as a number from 1 to 10 and the threshold is defined as a number from 1 to 100.

15. The method of claim 1, wherein a position of a user in a user interface results in the use of a lookup table to identify a matching programmable rule ID from a rules repository, where the rule ID is used to control a contextual relevance algorithm that will be applied in the user interface.

16. The method of claim 1, wherein a user repository containing the library of information may reside fully in a centralized database or at a customer location.

17. The method of claim 16, wherein the user repository, if located at a customer location, is synchronized with the centralized database and its user records on an initial and ongoing basis.

18. The method of claim 4, further comprising generating a relevance value for each match by using the relative weight of each matching metadata field that match the document of interest for matching.

19. The method of claim 4, further comprising generating a relevance value for each match by using the number of values that match the document of interest for matching.

20. The method of claim 4, further comprising generating a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching.

21. The method of claim 18, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

22. The method of claim 19, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

23. The method of claim 20, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

24. A system comprising:

a database to maintain a library of information; and

a contextual relevance engine to read metadata tagged information about a user'"'"'s work profile and working context and content within the library of information, match the metadata tagged information about content within the library of information to metadata tagged information about one or more of the user'"'"'s work profile, the user'"'"'s working context, or other content within the library of information using one or more programmable rules, and deliver to the user a rank ordered set of the most contextually relevant content from the library of information based on the matching of the metadata tagged information, wherein the one or more programmable rules each have a unique identifier (rule ID), a name, and a set of clauses defining the operations of that rule.

25. The system of claim 24, wherein each clause of the set of clauses comprises two or more of:

a name identifying the metadata field to be matched or operator to be performed;

a weight to be applied to the metadata field if a match is found;

an operator to act on the metadata field to be matched; and

a value for a field representing context.

26. The system of claim 25, wherein a clause comprises all of a name, a weight, an operator, and a value.

27. The system of claim 25, wherein the clause comprises a value forming a portion of a Boolean expression, and a name representing an operator used to combine all of the clauses.

28. The system of claim 27, wherein the operator comprises a Boolean operator.

29. The system of claim 27, wherein the operator comprises a “

matches all”

match that compares two or more values and excludes a document associated with the value if non-matching values are found.

30. The system of claim 27, wherein the operator comprises a “

matches”

match that compares two or more values, accumulates a score for matches Found, and excludes a document associated with the values if no matches are found.

31. The system of claim 27, wherein the operator comprises an “

optional”

match that compares two or more values and accumulate a score for matches found.

32. The system of claim 27, wherein the operator comprises an “

optional text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions and accumulates a score for matches found.

33. The system of claim 27, wherein the operator comprises a “

matches text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions, accumulates a score for matches found, and excludes a document associated with the values if no matches are found.

34. The system of claim 27, wherein the contextual relevance engine further:

generate a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching;

if the relevance value for a document exceeds a pre-defined threshold, deems the document to be relevant and includes the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deems the document to be irrelevant.

35. The system of claim 34, wherein the contextual relevance engine further sorts documents deemed to be relevant, by relevance, from most relevant to least relevant, and presenting, for display to the user, only a top tier of documents deemed to be most relevant.

36. The system of claim 35, wherein any document that matches itself in the list of relevant documents is discarded and is not presented for display to the user.

37. The system of claim 34, wherein each weight is defined as a number from Ito 10 and the threshold is defined as a number from Ito 100.

38. The system of claim 24, wherein a position of a user in a user interface results in the use of a lookup table to identify a matching programmable rule ID from a rules repository, where the rule ED is used to control a contextual relevance algorithm that will be applied in the user interface.

39. The system of claim 24, wherein a user repository containing the library of information may reside fully in a centralized database or at a customer location.

40. The system of claim 39, wherein the user repository, if located at a customer location, is synchronized with the centralized database and its user records on an initial and ongoing basis.

41. The system of claim 27, further comprising generating a relevance value for each match by using the relative weight of each matching metadata field that match the document of interest for matching.

42. The system of claim 27, further comprising generating a relevance value for each match by using the number of values that match the document of interest for matching.

43. The system of claim 27, further comprising generating a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching.

44. The system of claim 41, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

45. The system of claim 42, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

46. The system of claim 43, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

47. A machine-readable storage medium having stored thereon data representing sequences of instruction, the sequences of instructions which, when executed by a processor, cause the processor to:

maintain a library of information;

read metadata tagged information about a user'"'"'s work profile and working context and content within the library of information;

match the metadata tagged information about content within the library of information to metadata tagged information about one or more of the user'"'"'s work profile, the user'"'"'s working context, or other content within the library of information using a one or more programmable rules; and

deliver to the user a rank ordered set of the most contextually relevant content from the library of information based on the matching of the metadata tagged information, wherein the one or more programmable rules each have a unique identifier (rule ID), a name, and a set of clauses defining the operations of that rule.

48. The machine-readable medium of claim 47, wherein each clause of the set of clauses comprises two or more of:

a name identifying the metadata field to be matched or operator to be performed;

a weight to be applied to the metadata field if a match is found;

an operator to act on the metadata field to be matched; and

a value for a field representing context.

49. The machine-readable medium of claim 48, wherein a clause comprises all of a name, a weight, an operator, and a value.

50. The machine-readable medium of claim 48, wherein the clause comprises a value forming a portion of a Boolean expression, and a name representing an operator used to combine all of the clauses.

51. The machine-readable medium of claim 50, wherein the operator comprises a Boolean operator.

52. The machine-readable medium of claim 50, wherein the operator comprises a “

matches all”

match that compares two or more values and excludes a document associated with the value if non-matching values are found.

53. The machine-readable medium of claim 50, wherein the operator comprises a “

matches”

match that compares two or more values, accumulates a score for matches found, and excludes a document associated with the values if no matches are found.

54. The machine-readable medium of claim 50, wherein the operator comprises an “

optional”

match that compares two or more values and accumulates a score for matches found.

55. The machine-readable medium of claim 50, wherein the operator comprises an “

optional text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions and accumulates a score for matches found.

56. The machine-readable medium of claim 50, wherein the operator comprises a “

matches text”

match that breaks text strings associated with a document into component words excluding basic words, conjunctions, and prepositions, accumulates a score for matches found, and excludes a document associated with the values if no matches are found.

57. The machine-readable medium of claim 50, wherein the instruction further cause the processor to:

generate a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching;

if the relevance value for a document exceeds a pre-defined threshold, deem the document to be relevant and include the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deem the document to be irrelevant.

58. The machine-readable medium of claim 57, wherein the instructions further cause the processor to sort documents deemed to be relevant, by relevance, from most relevant to least relevant, and presenting, for display to the user, only a top tier of documents deemed to be most relevant.

59. The machine-readable medium of claim 58, wherein any document that matches itself in the list of relevant documents is discarded and is not presented for display to the user.

60. The machine-readable storage medium of claim 57, wherein each weight is defined as a number from 1 to 10 and the threshold is defined as a number from 1 to 100.

61. The machine-readable medium of claim 47, wherein a position of a user in a user interface results in the use of a lookup table to identify a matching programmable rule ID from a rules repository, where the rule ID is used to control a contextual relevance algorithm that will be applied in the user interface.

62. The machine-readable medium of claim 47, wherein a user repository containing the library of information may reside fully in a centralized database or at a customer location.

63. The machine-readable medium of claim 62, wherein the user repository, if located at a customer location, is synchronized with the centralized database and its user records on an initial and ongoing basis.

64. The machine-readable medium of claim 50, further comprising generating a relevance value for each match by using the relative weight of each matching metadata field that match the document of interest for matching.

65. The machine-readable medium of claim 50, further comprising generating a relevance value for each match by using the number of values that match the document of interest for matching.

66. The machine-readable medium of claim 50, further comprising generating a relevance value for each match by combining the relative weight of each matching metadata field and the number of values that match the document of interest for matching.

67. The machine-readable medium of claim 64, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to he relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a predefined threshold, deeming the document to be irrelevant.

68. The machine-readable medium of claim 65, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

69. The machine-readable medium of claim 66, further comprising:

if the relevance value for a document exceeds a pre-defined threshold, deeming the document to be relevant and including the document in a list of relevant documents; and

if the relevance value for a document does not exceed a pre-defined threshold, deeming the document to be irrelevant.

70. The machine-readable storage medium claim 47, wherein the contextually relevant content comprise office documents Microsoft Word documents, spreadsheets or presentations.

71. The machine-readable storage medium claim 47, wherein the contextually relevant content comprise rich media content text, images, audio, video and interactive media.