LINKAGE INFORMATION OUTPUT APPARATUS, LINKAGE INFORMATION OUTPUT METHOD AND COMPUTER-READABLE RECORDING MEDIUM

• US 20130007021A1
• Filed: 12/28/2010
• Published: 01/03/2013
• Est. Priority Date: 03/12/2010
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Mizuguchi, Hironori, Kusumura, Yukitaka, Kusui, Dai, Muraoka, Yusuke
• Granted Patent: US 9,152,696 B2
• Time in Patent Office: Days
• US Class Current: 707/750
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: G06F 16/9558 Details of hyperlinks; Management of linked annotations

Abstract

A linkage information output apparatus includes: a linkage information retrieval unit for acquiring, upon receiving source information, destination information linked with the source information, a frequency of occurrence of the source information, a frequency of occurrence of linked each of the destination information, and a frequency of occurrence of a link of the source information and each of the destination information from a linkage information accumulation unit; a recognition degree calculation unit calculating, based on each acquired frequency of occurrence, a recognition degree of the source information, a recognition degree of each acquired destination information, and a recognition degree of each link; and a high interest information narrowing unit selecting destination information to output from among each destination information based on a combination of two or more among a recognition degree of the source information, a recognition degree of the destination information, and a recognition degree of the link.

Claims

1-15. -15. (canceled)

16. A linkage information output apparatus for outputting information linked with source information comprising:

a linkage information retrieval unit which receives said source information, and acquires a plurality of destination information linked with said source information, a frequency of occurrence of said source information, a frequency of occurrence of linked each said destination information, and a frequency of occurrence of a link of said source information and said each destination information, from a linkage information accumulation unit provided inside or outside said linkage information output apparatus;

a recognition degree calculation unit which, based on acquired each said frequency of occurrence, calculates a recognition degree of said source information, a recognition degree of acquired said each destination information, and a recognition degree of each said link; and

a high interest information narrowing unit which selects one or more of said destination information from among said each destination information based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information and a recognition degree of said link, and outputs at least selected said destination information.

17. The linkage information output apparatus according to claim 16, wherein

said high interest information narrowing unit selects any two among said source information, said destination information, and said link, and selects, from among said each destination information, said destination information where a recognition degree of one of the selected ones becomes higher than a prescribed threshold value and a recognition degree of the other becomes lower than a prescribed threshold value.

18. The linkage information output apparatus according to claim 16, wherein

said high interest information narrowing unit selects, from among said each destination information, said destination information where at least one of a recognition degree of said source information and a recognition degree of said destination information becomes higher than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value.

19. The linkage information output apparatus according to claim 17, wherein

said high interest information narrowing unit, further, from among selected said destination information, selects destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and where a recognition degree of said link becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level.

20. The linkage information output apparatus according to claim 16, wherein

said recognition degree calculation unit, with respect to object information for which said recognition degree is to be calculated among said source information, acquired said each destination information and said each link, acquires a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and estimates an appearance probability distribution from acquired said total of frequencies of occurrence, and calculates an appearance probability of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

21. The linkage information output apparatus according to claim 16, wherein

said recognition degree calculation unit, with respect to an object information where said recognition degree is to be calculated among said source information, acquired said each destination information, and said link, acquires a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and calculates an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculates, as a recognition degree of said object information, a difference between said calculated average frequency of occurrence and the frequency of occurrence of said object information.

22. The linkage information output apparatus according to claim 16, wherein

said linkage information retrieval unit, further, acquires category information of said source information, and category information of linked said each destination information from said linkage information accumulation unit, and

said recognition degree calculation unit, from a total of frequencies of occurrence of links between categories of said source information and said destination information corresponding to an object link which is said link of which said recognition degree is to be calculated, estimates an appearance probability distribution, and using said estimated appearance probability distribution, calculates an appearance probability of the link between said categories as a recognition degree of said object link.

23. The linkage information output apparatus according to claim 22, wherein

said recognition degree calculation unit, in the case where at least one of said source information and said destination information which correspond to said object link belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degrees calculated with respect to each said category is made to be a recognition degree of said object link.

24. The linkage information output apparatus according to claim 16, wherein

said linkage information retrieval unit, further, acquires category information of said source information, and category information of linked said each destination information from said linkage information accumulation unit, and

said recognition degree calculation unit, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquires a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and estimates an appearance probability distribution from acquired said total of frequencies of occurrence, and calculates an appearance probability of the category of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

25. The linkage information output apparatus according to claim 16, wherein

said linkage information retrieval unit, further, acquires category information of said source information, and category information of linked said each destination information from said linkage information accumulation unit, and

said recognition degree calculation unit, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquires a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and calculates an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculates a difference between the calculated average frequency of occurrence and the frequency of occurrence of the category of said object information as a recognition degree of said object information.

26. The linkage information output apparatus according to claim 24, wherein

said recognition degree calculation unit, in the case where said object information belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degree calculated with respect to each said category is made to be a recognition degree of said object information.

27. The linkage information output apparatus according to claim 16, wherein

said linkage information retrieval unit, on receiving said source information, from said linkage information accumulation unit, acquires:

a plurality of destination information linked with said source information;

a relation content of said source information and each said destination information;

a frequency of occurrence of said source information;

a frequency of occurrence of said each destination information;

a frequency of occurrence of a link of said source information and said each destination information; and

a frequency of occurrence of each said relation content, and

said recognition degree calculation unit, based on acquired each said frequency of occurrence, calculates;

a recognition degree of said source information;

a recognition degree of acquired said each destination information;

a recognition degree of each said link; and

a recognition degree of acquired said each relation content, and

said high interest information narrowing unit, based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information, a recognition degree of said link and a recognition degree of said relation content, selects one or more of said destination information from said each destination information, and outputs at least selected said destination information.

28. The linkage information output apparatus according to claim 27, wherein

said high interest information narrowing unit selects, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become higher than a prescribed threshold value, and a recognition degree of said relation content becomes lower than a prescribed threshold value, and

selects, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value.

29. The linkage information output apparatus according to claim 28, wherein

said high interest information narrowing unit, further, selects, from among said each destination information, said destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, and

selects, from among said each destination information, said destination information where a recognition degree of one of said source information and said destination information become higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value, and

said high interest information narrowing unit, selects, from among selected said destination information, said destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level among said selected destination information, and selects said destination information where a recognition degree of one of said source information and said destination information becomes higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, a recognition degree of said link becomes lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value, as destination information where unpredictability is high in a middle level among selected said destination information.

30. The linkage information output apparatus according to claim 17, wherein

said linkage information retrieval unit, further, acquires category information of said source information, and category information of linked said each destination information from said linkage information accumulation unit, and

said recognition degree calculation unit, with respect to links between categories of each of said source information and said destination information which correspond to said relation content of which said recognition degree is to be calculated, calculates an appearance probability of said relation content based on a total of frequencies of occurrence of links between said categories and frequencies of occurrence of said relation content, and said calculated appearance probability is made to be a recognition degree of said relation content.

31. A linkage information output method in a linkage information output apparatus for outputting information linked with source information, the method comprising the steps of:

acquiring, on receiving said source information, a plurality of destination information linked with said source information, a frequency of occurrence of said source information, a frequency of occurrence of linked each said destination information, and a frequency of occurrence of a link of said source information and said each destination information, from a linkage information accumulation unit provided inside or outside said linkage information output apparatus;

calculating, based on acquired each said frequency of occurrence, a recognition degree of said source information, a recognition degree of acquired said each destination information, a recognition degree of each said link; and

selecting one or more of said destination information from among said each destination information based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information and a recognition degree of said link, and outputting at least selected said destination information.

32. The linkage information output method according to claim 31, wherein

in the step for selecting said destination information, selected are any two among said source information, said destination information, and said link, and selected from among said each destination information is said destination information where a recognition degree of one of the selected ones becomes higher than a prescribed threshold value and a recognition degree of the other becomes lower than a prescribed threshold value.

33. The linkage information output method according to claim 31, wherein

in the step for selecting said destination information, selected is, from among said each destination information, said destination information where at least one of a recognition degree of said source information or a recognition degree of said destination information becomes higher than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value.

34. The linkage information output method according to claim 32, wherein

in the step for selecting said destination information, further, from among selected said destination information, selected is destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and where a recognition degree of said link becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level.

35. The linkage information output method according to claim 31, wherein

in the step for calculating said recognition degree, with respect to object information for which said recognition degree is to be calculated among said source information and acquired said each destination information and said each link, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and estimated is an appearance probability distribution from acquired said total of frequencies of occurrence, and calculated is an appearance probability of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

36. The linkage information output method according to claim 31, wherein

in the step for calculating said recognition degree, with respect to an object information where said recognition degree is to be calculated among said source information, acquired said each destination information, and said link, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and calculated is an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculated is, as a recognition degree of said object information, a difference between said calculated average frequency of occurrence and the frequency of occurrence of said object information.

37. The linkage information output method according to claim 31, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, from a total of frequencies of occurrence of links between categories of said source information and said destination information corresponding to an object link which is said link of which a recognition degree is to be calculated, estimated is an appearance probability distribution, and using said estimated appearance probability distribution, calculated is, as a recognition degree of said object link, an appearance probability of a link between said categories.

38. The linkage information output method according to claim 37, wherein

in the step for calculating said recognition degree, in the case where at least one of said source information and said destination information which correspond to said object link belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degrees calculated with respect to each said category is made to be a recognition degree of said object link.

39. The linkage information output method according to claim 31, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all the links accumulated in said linkage information accumulation unit, and estimated is an appearance probability distribution from acquired said total of frequencies of occurrence, and calculated is an appearance probability of a category of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

40. The linkage information output method according to claim 31, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all the links accumulated in said linkage information accumulation unit, and calculated is an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculated is a difference between the calculated average frequency of occurrence and the frequency of occurrence of the category of said object information as a recognition degree of said object information.

41. The linkage information output method according to claim 39, wherein

in the step for calculating said recognition degree, in the case where said object information belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degree calculated with respect to each said category is made to be a recognition degree of said object information.

42. The linkage information output method according to claim 31, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, on receiving said source information, from said linkage information accumulation unit, acquired are a plurality of destination information linked with said source information, a relation content of said source information and each said destination information, a frequency of occurrence of said source information, and a frequency of occurrence of said each destination information, a frequency of occurrence of a link of said source information and said each destination information, and a frequency of occurrence of each said relation content, and

in the step for calculating said recognition degree, based on acquired each said frequency of occurrence, calculated are a recognition degree of said source information, a recognition degree of acquired said each destination information, a recognition degree of each said link, and a recognition degree of acquired said each relation content, and

in the step for selecting said destination information, based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information, a recognition degree of said link and a recognition degree of said relation content, selected are one or more of said destination information from said each destination information, and outputted is at least selected said destination information.

43. The linkage information output method according to claim 42, wherein

in the step for selecting said destination information, selected is, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become higher than a prescribed threshold value, and a recognition degree of said relation content becomes lower than a prescribed threshold value, and selected is, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value.

44. The linkage information output method according to claim 43, wherein

in the step for selecting said destination information, further, selected is, from among said each destination information, said destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level among said selected destination information, and

selected is, from among said each destination information, said destination information where a recognition degree of one of a recognition degree of said source information and said destination information becomes higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value, and

from among selected said destination information, selected is destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level among said selected destination information, and selected is said destination information where a recognition degree of one of a recognition degree of said source information and said destination information becomes higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, a recognition degree of said link becomes lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value, as destination information where unpredictability is high in a middle level among selected said destination information.

45. The linkage information output method according to claim 42, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to links between categories of each of said source information and said destination information which correspond to said relation content of which said recognition degree is to be calculated, calculated is an appearance probability of said relation content based on a total of frequencies of occurrence of links between said categories and frequencies of occurrence of said relation content, and said calculated appearance probability is made to be a recognition degree of said relation content.

46. A computer-readable recording medium having a linkage information output program used in a linkage information output apparatus for outputting information linked with source information recorded, the linkage information output program making a computer execute the steps of:

acquiring, on receiving said source information, a plurality of destination information linked with said source information, a frequency of occurrence of said source information, a frequency of occurrence of linked each said destination information, and a frequency of occurrence of a link of said source information and said each destination information, from a linkage information accumulation unit provided inside or outside said linkage information output apparatus;

calculating, based on acquired each said frequency of occurrence, a recognition degree of said source information, a recognition degree of acquired said each destination information, a recognition degree of each said link; and

selecting one or more of said destination information from among said each destination information based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information and a recognition degree of said link, and outputting at least selected said destination information.

47. The computer-readable recording medium according to claim 46, wherein

in the step for selecting said destination information, selected are any two among said source information, said destination information, and said link, and selected from among said each destination information is said destination information where a recognition degree of one of the selected ones becomes higher than a prescribed threshold value and a recognition degree of the other becomes lower than a prescribed threshold value.

48. The computer-readable recording medium according to claim 46, wherein

in the step for selecting said destination information, selected is, from among said each destination information, said destination information where at least one of a recognition degree of said source information or a recognition degree of said destination information becomes higher than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value.

49. The computer-readable recording medium according to claim 47, wherein

in the step for selecting said destination information, further, from among selected said destination information, selected is destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and where a recognition degree of said link becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level.

50. The computer-readable recording medium according to claim 46, wherein

in the step for calculating said recognition degree, with respect to object information for which said recognition degree is to be calculated among said source information and acquired said each destination information and said each link, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and estimated is an appearance probability distribution from acquired said total of frequencies of occurrence, and calculated is an appearance probability of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

51. The computer-readable recording medium according to claim 46, wherein

in the step for calculating said recognition degree, with respect to object information for which said recognition degree is to be calculated among said source information, acquired said each destination information, and said each link, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of all of source information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all of destination information accumulated in said linkage information accumulation unit and a total of frequencies of occurrence of all the links accumulated in said linkage information accumulation unit, and calculated is an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculated is, as a recognition degree of said object information, a difference between said calculated average frequency of occurrence and the frequency of occurrence of said object information.

52. The computer-readable recording medium according to claim 46, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, from a total of frequencies of occurrence of links between categories of said source information and said destination information corresponding to an object link which is said link of which a recognition degree is to be calculated, estimated is an appearance probability distribution, and using said estimated appearance probability distribution, calculated is, as a recognition degree of said object link, an appearance probability of a link between said categories.

53. The computer-readable recording medium according to claim 52, wherein

in the step for calculating said recognition degree, in the case where at least one of said source information and said destination information which correspond to said object link belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degrees calculated with respect to each said category is made to be a recognition degree of said object link.

54. The computer-readable recording medium according to claim 46, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all the links accumulated in said linkage information accumulation unit, and estimated is an appearance probability distribution from acquired said total of frequencies of occurrence, and calculated is an appearance probability of a category of said object information as a recognition degree of said object information using said estimated appearance probability distribution.

55. The computer-readable recording medium according to claim 46, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to object information of which said recognition degree is to be calculated among said source information and acquired said each destination information, acquired are a total of frequencies of occurrence corresponding to said object information among a total of frequencies of occurrence of categories of all of source information accumulated in said linkage information accumulation unit, a total of frequencies of occurrence of categories of all of destination information accumulated in said linkage information accumulation unit, and a total of frequencies of occurrence of categories of all the links accumulated in said linkage information accumulation unit, and calculated is an average frequency of occurrence from acquired said total of frequencies of occurrence, and calculated is a difference between the calculated average frequency of occurrence and the frequency of occurrence of the category of said object information as a recognition degree of said object information.

56. The computer-readable recording medium according to claim 54, wherein

in the step for calculating said recognition degree, in the case where said object information belongs to a plurality of categories, an average value, the maximum or the minimum of said recognition degree calculated with respect to each said category is made to be a recognition degree of said object information.

57. The computer-readable recording medium according to claim 46, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, on receiving said source information, from said linkage information accumulation unit, acquired are a plurality of destination information linked with said source information, a relation content of said source information and each said destination information, a frequency of occurrence of said source information, and a frequency of occurrence of said each destination information, a frequency of occurrence of a link of said source information and said each destination information, and a frequency of occurrence of each said relation content, and

in the step for calculating said recognition degree, based on acquired each said frequency of occurrence, calculated are a recognition degree of said source information, a recognition degree of acquired said each destination information, a recognition degree of each said link, and a recognition degree of acquired said each relation content, and

in the step for selecting said destination information, based on a combination of two or more among a recognition degree of said source information, a recognition degree of said destination information, a recognition degree of said link and a recognition degree of said relation content, selected are one or more of said destination information from said each destination information, and outputted is at least selected said destination information.

58. The computer-readable recording medium according to claim 57, wherein

in the step for selecting said destination information, selected is, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become higher than a prescribed threshold value, and a recognition degree of said relation content becomes lower than a prescribed threshold value, and selected is, from among said each destination information, said destination information where a recognition degree of said source information, a recognition degree of said destination information, and a recognition degree of said link become lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value.

59. The computer-readable recording medium according to claim 58, wherein

in the step for selecting said destination information, further, selected is, from among said each destination information, said destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level among said selected destination information, and

selected is, from among said each destination information, said destination information where a recognition degree of one of a recognition degree of said source information and said destination information becomes higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, and a recognition degree of said link becomes lower than a prescribed threshold value, and

from among selected said destination information, selected is destination information where a recognition degree of said source information and a recognition degree of said destination information become higher than a prescribed threshold value, and at least one of a recognition degree of said link and a recognition degree of said relation content becomes lower than a prescribed threshold value, as destination information where unpredictability is high in a high level among said selected destination information, and selected is said destination information where a recognition degree of one of a recognition degree of said source information and said destination information becomes higher than a prescribed threshold value, and a recognition degree of the other becomes lower than a prescribed threshold value, a recognition degree of said link becomes lower than a prescribed threshold value, and a recognition degree of said relation content becomes higher than a prescribed threshold value, as destination information where unpredictability is high in a middle level among selected said destination information.

60. The computer-readable recording medium according to claim 57, wherein

in the step for acquiring said each destination information and said each frequency of occurrence, further, acquired are category information of said source information and category information of linked said each destination information from said linkage information accumulation unit, and

in the step for calculating said recognition degree, with respect to links between categories of each of said source information and said destination information which correspond to said relation content of which said recognition degree is to be calculated, calculated is an appearance probability of said relation content based on a total of frequencies of occurrence of links between said categories and frequencies of occurrence of said relation content, and said calculated appearance probability is made to be a recognition degree of said relation content.