Method and apparatus for assigning a confidence level to a term within a user knowledge profile

• US 6,832,224 B2
• Filed: 04/01/2002
• Issued: 12/14/2004
• Est. Priority Date: 09/18/1998
• Status: Expired due to Term

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Tacit Software, Inc. (Oracle Corporation)
• Inventors: Gilmour, David L.
• Primary Examiner(s): Alam, Shahid
• Assistant Examiner(s): Fleurantin, Jean Bolte
• Application Number:
  US10/114,179
• Publication Number:
  US 20020165861A1
• Time in Patent Office: 988 Days
• Field of Search: 707/3-10, 707/100-104, 707/1, 715/500, 729/206-207
• US Class Current: 1/1
• CPC Class Codes: G06F 16/313 Selection or weighting of terms for indexing
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A method of assigning a confidence level to a term within an electronic document, such as an e-mail, includes the step of firstly determining a quantitative indicator in the exemplary form of an occurrence value, based on the number of occurrences of a particular term within an electronic document, and associating the occurrence term within the relevant term. Thereafter, a qualitative indicator, based on a quality of the term, is determined. For example, the qualitative indicator may be determined utilizing the parts of speech of words comprising the term. A confidence level value, which may be utilized to indicate a relative importance of the term in describing a user knowledge base, is then generated utilizing the quantitative and qualitative indicators.

Claims

1. A computer implemented method comprising

dynamically discovering a term in a current electronic document without reference to a pre-defined value and without reference to a pre-defined category, the current electronic document being received from an entity;

deriving a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term; and

assigning the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity.

2. The method of claim 1 further comprising:

adding the term to the profile if the term first occurs in the current electronic document.

3. The method of claim 1, wherein deriving the current confidence level comprises:

calculating a quantitative indicator for the term in a dynamic set of electronic documents;

determining a term weight based on characteristics of the term; and

generating a relevancy indicator from the quantitative indicator and the term weight.

4. The method of claim 3, wherein deriving the current confidence level further comprises:

scaling the relevancy indicator.

5. The method of claim 4, wherein the relevancy indicator is scaled based on a measure of a size of the dynamic set of electronic documents.

6. The method of claim 3, wherein the characteristics of the term are selected from the group consisting of a number of words, a part of speech, and a type of grammatical structure.

7. The method of claim 1, wherein the current confidence level is based on a dynamic set of electronic document and further comprising:

excluding electronic documents from the dynamic set of electronic documents in accordance with expiration criteria associated with the dynamic set.

8. The method of claim 7, wherein the expiration criteria is selected from the group consisting of a time period and a number of documents.

9. The method of claim 1 further comprising:

decaying a highest confidence level recorded for the term by a decay value if the current confidence level is less than the highest confidence level and the highest confidence level occurred outside a document window.

10. The method of claim 9, wherein the document window is selected from the group consisting of a time period and a number of documents.

11. The method of claim 10 further comprising:

setting the highest confidence level to the current confidence level if the highest confidence level is less than the current confidence level.

12. The method of claim 1 wherein the entity is a person.

13. The method of claim 1 wherein the entity is a group of people.

14. A method comprising:

dynamically discovering a term in a current electronic document without reference to a pre-defined value and without reference to a pre-defined category, the current electronic document being received from an entity;

deriving a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term; and

assigning the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity, wherein deriving the current confidence level comprises;

calculating a quantitative indicator for the term in a dynamic set of electronic documents, wherein calculating the quantitative indicator comprises determining a binding strength for the term in the current document, the binding strength representing a measure of importance of the term in the current document, deriving an adjusted count value for the term in the current document from the binding strength and the term weight, and summing all adjusted count values for the term in the dynamic set of electronic documents;

determining a term weight based on characteristics of the term; and

generating a relevancy indicator from the quantitative indicator and the term weight.

15. The method of claim 14, wherein the binding strength is based on a count value and a density value, the count value specifying a number of occurrences of the term in the current electronic document and the density value relating the count value to a total number of terms in the current electronic document.

16. The method of claim 15, wherein the binding strength is based on the density value when the density value is above a predetermined value and on the count value when the density value is below the predetermined value.

17. A computer-readable medium having executable instructions to cause a computer to perform a method comprising:

dynamically discovering a term in a current electronic document without reference to a set of pre-defined values and without reference to a set of pre-defined categories, the current electronic document being received from an entity;

deriving a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term; and

assigning the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity.

18. The computer-readable medium of claim 17, wherein the method further comprises:

adding the term to the profile if the term first occurs in the current electronic document.

19. The computer-readable medium of claim 17, wherein deriving the current confidence level comprises:

calculating a quantitative indicator for the term in a dynamic set of electronic documents;

determining a term weight based on characteristics of the term; and

generating a relevancy indicator from the quantitative indicator and the term weight.

20. The computer-readable medium of claim 19, wherein deriving the current confidence level further comprises:

scaling the relevancy indicator.

21. The computer-readable medium of claim 20, wherein the relevancy indicator is scaled based on a measure of a size of the dynamic set of electronic documents.

22. The computer-readable medium of claim 19, wherein the characteristics of the term are selected from the group consisting of a number of words, a part of speech, and a type of grammatical structure.

23. The computer-readable medium of claim 17, wherein the current confidence level is based on a dynamic set of electronic document and the method further comprises:

excluding electronic documents from the dynamic set of electronic documents in accordance with expiration criteria associated with the dynamic set.

24. The computer-readable medium of claim 23, wherein the expiration criteria is selected from the group consisting of a time period and a number of documents.

25. The computer-readable medium of claim 17, wherein the method further comprises:

decaying a highest confidence level recorded for the term by a decay value if the current confidence level is less than the highest confidence level and the highest confidence level occurred outside a document window.

26. The computer-readable medium of claim 25, wherein the document window is selected from the group consisting of a time period and a number of documents.

27. The computer-readable medium of claim 25, wherein the method further comprises:

setting the highest confidence level to the current confidence level if the highest confidence level is less than the current confidence level.

28. The computer-readable medium of claim 17 wherein the entity is a person.

29. The computer-readable medium of claim 17 wherein the entity is a group of people.

30. A computer-readable medium having executable instructions to cause a computer to perform a method comprising:

dynamically discovering a term in a current electronic document without reference to a set of pre-defined values and without reference to a set of pre-defined categories, the current electronic document being received from an entity;

deriving a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term; and

assigning the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity, wherein deriving the current confidence level comprises;

calculating a quantitative indicator for the term in a dynamic set of electronic documents, wherein calculating the quantitative indicator comprises determining a binding strength for the term in the current document, the binding strength representing a measure of importance of the term in the current document, deriving an adjusted count value for the term in the current document from the binding strength and the term weight, and summing all adjusted count values for the term in the dynamic set of electronic documents;

determining a term weight based on characteristics of the term; and

generating a relevancy indicator from the quantitative indicator and the term weight.

31. The computer-readable medium of claim 30, wherein the binding strength is based on a count value and a density value, the count value specifying a number of occurrences of the term in the current electronic document and the density value relating the count value to a total number of terms in the current electronic document.

32. The computer-readable medium of claim 31, wherein the binding strength is based on the density value when the density value is above a predetermined value and on the count value when the density value is below the predetermined value.

33. A computer system comprising:

a processor; and

a memory coupled to the processor and having a confidence value process to be executed by the processor to cause the processor to dynamically discover a term in a current electronic document without reference to a set of pre-defined values and without reference to a set of pre-defined categories, the current electronic document being received from an entity, to derive a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term, and to assign the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity.

34. The computer system of claim 33, wherein the confidence value process further causes the processor to add the term to the profile if the term first occurs in the current electronic document.

35. The computer system of claim 33, wherein the confidence value process further causes the processor, when deriving the current confidence level, to calculate a quantitative indicator for the term in a dynamic set of electronic documents, to determine a term weight based on characteristics of the term, and to generate a relevancy indicator from the quantitative indicator and the term weight.

36. The computer system of claim 35, wherein the confidence value process further causes the processor, when deriving the current confidence level, to scale the relevancy indicator.

37. The computer system of claim 36, wherein the confidence value process further causes the processor to scale the relevancy indicator based on a measure of a size of the dynamic set of electronic documents.

38. The computer system of claim 35, wherein the confidence value process further causes the processor to select the characteristics of the term from the group consisting of a number of words, a part of speech, and a type of grammatical structure.

39. The computer system of claim 33, wherein the confidence value process further causes the processor to base the current confidence level on a dynamic set of electronic document and to exclude electronic documents from the dynamic set of electronic documents in accordance with expiration criteria associated with the dynamic set.

40. The computer system of claim 39, wherein the confidence value process further causes the processor to select the expiration criteria from the group consisting of a time period and a number of documents.

41. The computer system of claim 33, wherein the confidence value process further causes the processor to decay a highest confidence level recorded for the term by a decay value if the current confidence level is less than the highest confidence level and the highest confidence level occurred outside a document window.

42. The computer system of claim 41, wherein the confidence value process further causes the processor to select the document window from the group consisting of a time period and a number of documents.

43. The computer system of claim 41, wherein the confidence value process further causes the processor to set the highest confidence level to the current confidence level if the highest confidence level is less than the current confidence level.

44. The computer system of claim 33 wherein the entity is a person.

45. The computer system of claim 33 wherein the entity is a group of people.

46. A computer system comprising:

a processor; and

a memory coupled to the processor and having a confidence value process to be executed by the processor to cause the processor to dynamically discover a term in a current electronic document without reference to a set of pre-defined values and without reference to a set of pre-defined categories, the current electronic document being received from an entity;

to derive a current confidence level for the term to measure a current strength of a dynamic association between the entity and information represented by the term, when deriving the current confidence level, to calculate a quantitative indicator for the term in a dynamic set of electronic documents, to determine a term weight based on characteristics of the term, and to generate a relevancy indicator from the quantitative indicator and the term weight; and

to assign the current confidence level to the term in a profile for the entity, the profile comprising terms representing an information focus of the entity, wherein the confidence value process further causes the processor, when calculating the quantitative indicator, to determine a binding strength for the term in the current document, the binding strength representing a measure of importance of the term in the current document, to derive an adjusted count value for the term in the current document from the binding strength and the term weight, and to sum all adjusted count values for the term in the dynamic set of electronic documents.

47. The computer system of claim 46, wherein the confidence value process further causes the processor to base the binding strength on a count value and a density value, the count value specifying a number of occurrences of the term in the current electronic document and the density value relating the count value to a total number of terms in the current electronic document.

48. The computer system of claim 47, wherein the confidence value process further causes the processor to base the binding strength on the density value when the density value is above a predetermined value and on the count value when the density value is below the predetermined value.