System and method for creating, editing, and using multi-relational ontologies

• US 20060053172A1
• Filed: 05/05/2005
• Published: 03/09/2006
• Est. Priority Date: 09/03/2004
• Status: Abandoned Application

• Current Assignee: BioWisdom Ltd. (Instem Plc)
• Original Assignee: BioWisdom Ltd. (Instem Plc)
• Inventors: Davis, Benjamin, Baxter, Gordon Smith, Eldridge, Matthew David, Gardner, Stephen Philip, Hill, Robin Duncan, McMenamin, Conor, Beaumont, Simon Edwin, Barnes, Julie Christine
• Application Number:
  US11/122,066
• Publication Number:
  US 20060053172A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation

Abstract

The invention relates to a system and method for creating, editing, using one or more multi-relational ontologies and applying knowledge contained in one or more ontologies to one or more applications. The system may extract concepts, relationships, assertions and other information from a domain specific corpus of documents. The system may also create rule-based assertions that are then curated and published as one or more ontologies that can be used in various ways.

Claims

1. A computer-implemented system for creating one or more domain specific multi-relational ontologies, comprising:

means for creating an upper ontology that specifies, for a specific domain, a set of concept types and relationship types, property types associated with concept types, and a hierarchy of concept types and relationship types;

means for storing the upper ontology;

means for selecting a corpus of documents from one or more data sources, at least some of the documents being related to the specific domain;

a rules engine for creating and storing predefined rules relating to the creation of assertions, where assertions include a first concept, a second concept and a relationship between the first and second concept;

an extraction module for;

i) extracting from the corpus of documents, in accordance with the rules, concepts and relationships between concepts to form rules-based assertions and for specifying for the extracted concepts;

a) the type of concept;

b) the type of relationship with other concepts;

c) properties of the concept;

where the concept types and relationship types are constrained by the upper ontology; and

ii) associating evidence information with the rule based assertions;

an editor module including an interface for enabling a curator to;

i) view, edit and validate at least some of the rules-based assertions; and

ii) create new assertions;

to form a set of reified assertions; and

means for storing the reified assertions and evidence information in a database as a domain specific ontology.

2. The system of claim 1, wherein the upper ontology defines each concept type and for each concept type, the relationship types and property types that may exist for that concept type.

3. The system of claim 1, wherein the upper ontology further comprises data regarding data sources.

4. The system of claim 1, wherein the upper ontology further comprises means for defining each data source in terms of what it contains and what can be extracted from the data source.

5. The system of claim 1, wherein the upper ontology further comprises means for storing curator information.

6. The system of claim 1, wherein the upper ontology further comprises means for storing curator activity data.

7. The system of claim 1, wherein the upper ontology further comprises versioning information for the domain specific ontology.

8. The system of claim 1, wherein the upper ontology further comprises time stamp information regarding assertions.

9. The system of claim 1, wherein the upper ontology further comprises time stamp information including when assertions were deemed true, when assertions were disproved, and when assertions were entered into the domain specific ontology.

10. The system of claim 1, further comprising means for specifying a corpus for a specific domain, including selecting from:

i) one or more structured data sources;

ii) one or more unstructured data sources;

iii) one or more semi-structured data sources;

iv) one or more web searches; and

v) one or more searches of locally or remotely available information sources.

11. The system of claim 1, wherein at least one of the one or more data sources comprises a website, and wherein the means for extracting further comprises a web-crawler for extracting the constituent concepts and relationships from the at least one website.

12. The system of claim 1, wherein the extraction module comprises an automated rules-based text mining module.

13. The system of claim 1, wherein the extraction module comprises means for associating evidence information with each assertion, where the evidence information includes one or more of:

i) at least one data source; and

ii) at least one document from with the data source.

14. The system of claim 1, wherein the extraction module comprises means for associating evidence information with each assertion, where the evidence information includes a link to the document evidencing the assertion.

15. The system of claim 1, wherein the extraction module comprises means for associating evidence information with each assertion, where the evidence information includes a link to a portion of the document evidencing the assertion.

16. The system of claim 1, wherein the extraction module comprises means for associating evidence information with each assertion, where the evidence information includes a link to a portion of a document evidencing the assertion, and further wherein one or more words evidencing the assertion are highlighted.

17. The system of claim 1, further comprising means for automatically semantically normalizing assertions.

18. The system of claim 1, wherein the editor module comprises a document viewer that enables a curator to view one or more documents associated with a rules-based assertion.

19. The system of claim 1, further comprising means for storing, for each assertion, one or more of:

at least one data source, the type of the at least one data source, the mechanism(s) by which the assertion was extracted, when the assertion was extracted, and what version of data the assertion was extracted from.

20. The system of claim 1, further comprising means for storing, with a reified assertion, which curator validated, edited, or created the assertion.

21. The system of claim 1, further comprising means for storing, with a reified assertion, a confidence level or other weighing factor associated with the assertion.

22. The system of claim 1, further comprising means for storing with a reified assertion, a classification of the assertion.

23. The system of claim 1, further comprising means for storing with a reified assertion, a classification of the assertion, wherein the classifications include one or more of:

i) a factual assertion;

ii) a strongly evidenced assertion; and

iii) a weakly evidenced assertion.

24. The system of claim 1, further comprising a publishing module for publishing an ontology to make the domain specific ontology accessible to one or more users.

25. The system of claim 1, further comprising a publishing module for publishing an ontology, including means for performing a schema conversion.

26. The system of claim 1, further comprising an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies.

27. The system of claim 1, further comprising an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies based on one or more of a user profile and user access rights.

28. The system of claim 1, further comprising an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies based on one or more of a user profile, wherein the user profile comprises user preferences regarding one or more of:

i) themes and types;

ii) data source or data source type; and

iii) curator information.

29. The system of claim 1, further comprising an export manager module for exporting one or more ontologies or portions of the one or more ontologies to another application for further use of the ontologies.

30. The system of claim 1, further comprising an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies including means for merging two more ontologies or portions of two or more ontologies.

31. The system of claim 1, further comprising means for generating custom ontologies based on one or more ontologies or portions of the one or more ontologies.

32. The system of claim 1, further comprising means for generating custom ontologies based on one or more ontologies or portions of the one or more ontologies and a set of filter criteria.

33. The system of claim 1, further comprising an alerts module, wherein the alerts module generates user alerts based on ontologies activity and user preferences.

34. The system of claim 1, further comprising a graphical user interface through which a user can view and interact with one or more ontologies.

35. The system of claim 1, further comprising a graphical user interface that enables a user to view ontology information, including a pane for displaying a hierarchical view of concepts within the domain specific ontology and a pane for displaying a multi-relational view of a network of concepts related to a selected concept.

36. A computer-implemented method for creating one or more domain specific ontologies, the method comprising the steps of:

creating an upper ontology that specifies, for a specific domain, a set of concept types, property types, and relationship types, and a hierarchy of concept types and relationship types;

storing the upper ontology;

selecting a corpus of documents from one or more data sources, at least some of the documents being relating to the specific domain;

providing a rules engine for creating and storing predefine rules relating to the creation of assertions, where assertions include a first concept, a second concept and a relationship between the first and second concept;

using an extraction module for;

iii) extracting from the corpus of documents, in accordance with the rules, concepts and relationships between concepts to form rules-based assertions and for specifying for the extracted concepts;

a) the type of concept;

b) the type of relationship with other concepts;

c) property types;

where the concept types, relationship types, and property types are constrained by the upper ontology; and

iv) associating evidence information with the extracted assertions;

using an editor module to enable a curator to perform the steps of;

i) viewing, editing and validating at least some of the rules-based assertions; and

ii) creating new assertions;

to form a set of reified assertions; and

storing the reified assertions and evidence information in a database as a domain specific ontology.

37. The method of claim 36, wherein the upper ontology defines each concept type and for each concept type, the relationship types and property types that may exist for that concept type.

38. The method of claim 36, wherein the upper ontology further comprises data regarding data sources.

39. The method of claim 36, wherein the upper ontology further comprises means for defining each data source in terms of what it contains and what can be extracted from the data source.

40. The method of claim 36, wherein the upper ontology further comprises means for storing curator information.

41. The method of claim 36, wherein the upper ontology further comprises means for storing curator activity data.

42. The method of claim 36, wherein the upper ontology further comprises versioning information for each ontology.

43. The method of claim 36, wherein the upper ontology further comprises time stamp information regarding assertions.

44. The method of claim 36, wherein the upper ontology further comprises time stamp information including when assertions were deemed true, when assertions were disproved and when assertions were entered into the ontology.

45. The method of claim 36, wherein the step of specifying a corpus for a specific domain, including selecting from:

i) one or more structured data sources;

ii) one or more unstructured data sources;

iii) one or more semi-structured data sources;

iv) one or more web searches; and

v) one or more searches of locally or remotely available information sources.

46. The method of claim 36, wherein at least one of the one or more data sources is a website, and wherein the step of extracting further comprises using a web-crawler for extracting the constituent concepts and relationships from the at least one website.

47. The method of claim 36, wherein the step of extracting comprises using an automated rules-based text mining module.

48. The method of claim 36, wherein the step of extracting comprises associating evidence information with each assertion, where the evidence information includes one or more of:

i) at least one data source; and

ii) at least one document from with the data source.

49. The method of claim 36, wherein the step of extracting comprises associating evidence information with each assertion, where the evidence information includes a link to the document evidencing the assertion.

50. The method of claim 36, wherein the step of extracting comprises means for associating evidence information with each assertion, where the evidence information includes a link to a portion of a document evidencing the assertion.

51. The method of claim 36, wherein the step of extracting comprises associating with each assertion evidence information, where the evidence information includes a link to a portion of a document evidencing the assertion, and further comprising the step of automatically highlighting one or more words evidencing the assertion.

52. The method of claim 36, further comprising the step of automatically semantically normalizing assertions.

53. The method of claim 36, wherein the step of using an editor module comprises using a document viewer to enable a curator to view one or more documents associated with a rules-based assertion.

54. The method of claim 36, further comprising the step of storing, for each assertion, one or more of:

at least one data source, the type of the at least one data source, the mechanism(s) by which the assertion was extracted, when the assertion was extracted, and what version of data the assertion was extracted from.

55. The method of claim 36, further comprising the step of storing, with a reified assertion, which curator validated, edited or created the assertion.

56. The method of claim 36, further comprising the step of storing, with a reified assertion, a confidence level associated with the assertion.

57. The method of claim 36, further comprising the step of storing with a reified assertion, a classification of the assertion.

58. The method of claim 36, further comprising the step of storing with a reified assertion, a classification of the assertion, wherein the classifications include one or more of:

i) a factual assertion;

ii) a strongly evidenced assertion; and

iii) a weakly evidenced assertion.

59. The method of claim 36, further comprising the step of publishing an ontology to make the ontology accessible to one or more users.

60. The method of claim 36, further comprising the step of publishing an ontology, including performing a schema conversion.

61. The method of claim 36, further comprising the step of using an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies.

62. The method of claim 36, further comprising the step of using an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies based on one or more of a user profile and user access rights.

63. The method of claim 36, further comprising the step of using an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies based on one or more of a user profile, wherein the user profile comprises user preferences regarding one or more of:

i) themes and types;

ii) data source or data source type; and

iii) curator information.

64. The method of claim 36, further comprising the step of using an export manager module for exporting one or more ontologies or portions of the one or more ontologies to another application for further use of the domain specific ontology.

65. The method of claim 36, further comprising the step of using an export manager module for controlling the export of one or more ontologies or portions of the one or more ontologies including means for merging two more ontologies or portions of two or more ontologies.

66. The method of claim 36, further comprising the step of generating custom ontologies based on one or more ontology or portions of the one or more ontologies.

67. The method of claim 36, further comprising the step of generating custom ontologies based on one or more ontologies or portions of the one or more ontologies and a set of filter criteria.

68. The method of claim 36, further comprising the step of generating user alerts based on ontology activity and user preferences.

69. The method of claim 36, further comprising the step of providing a graphical user interface through which a user can view and interact with one or more ontologies.

70. The method of claim 36, further comprising the step of providing a graphical user interface that enables a user to view ontology information, including a pane for displaying a hierarchical view of concepts within the ontology and a pane for displaying a multi-relational view of a network of concepts related to a selected concept.