System and method for generating and using user ontological models for natural language processing of user-provided text

• US 9,588,962 B2
• Filed: 03/20/2015
• Issued: 03/07/2017
• Est. Priority Date: 02/03/2015
• Status: Active Grant

• Current Assignee: ABBYY Development LLC
• Original Assignee: ABBYY InfoPoisk LLC
• Inventors: Bulgakov, Ilia, Yakovlev, Egor, Starostin, Anatoly
• Primary Examiner(s): He, Jialong
• Application Number:
  US14/663,797
• Publication Number:
  US 20160224542A1
• Time in Patent Office: 718 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/367 Ontology
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/237 Lexical tools
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

A method, system, and computer program product for generating and using a user ontological model for natural language processing of user-provided text, including receiving definitions of user ontological objects and generating user ontological models. A semantic-syntactic tree generated from user-provided text is analyzed. Information objects based on the user ontological objects are generated by the analysis.

Claims

1. A method comprising:

receiving, by an interface from a user, a definition of an instance of a concept, wherein the instance is specific to a particular field of knowledge of the user;

in response to receiving the definition, generating, by a hardware processor, a user ontological model based on the definition, wherein the user ontological model is created for the user and the particular field of knowledge of the user, wherein a main ontological model provided to the user does not comprise ontological objects specific to the particular field of knowledge, wherein the user ontological model and the main ontological model are for use in natural language processing of natural language text, wherein the user ontological model comprises an instance structure that is based on the definition, and wherein the instance structure has an identifier;

accessing a semantic-syntactic tree generated from a text provided by the user, wherein the semantic-syntactic tree comprises a plurality of connected nodes, and wherein at least one node among the plurality of connected nodes comprises a reference to the identifier of the instance structure;

identifying the instance structure in the user ontological model based on the reference to the identifier in the semantic-syntactic tree;

in response to identifying the instance structure, generating a first information object for the node in the semantic-syntactic tree based on the instance structure; and

performing, by the hardware processor, natural language processing of the text based on the first information object.

2. The method of claim 1, further comprising:

identifying a fragment in the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node;

determining, from the ontological rule and from the fragment, a property of the instance structure;

determining, from the ontological rule and from the fragment, a value of the property; and

assigning the value to a same property of the first information object as the property of the instance structure.

3. The method of claim 1, further comprising:

identifying a fragment of the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node; and

generating a second information object corresponding to the pattern of nodes and to the fragment.

4. The method of claim 3, further comprising:

determining, from the ontological rule and from the fragment, an ontological object property of an ontological object in the user ontological model or the main ontological model corresponding to the second information object; and

setting a value of a property of the second information object to a value of a same property as the first information object.

5. The method of claim 3, further comprising:

determining, from the ontological rule and from the fragment, a property of the instance structure; and

setting a value of a property of the first information object to a value of a same property as the second information object.

6. The method of claim 1, further comprising generating an object-property-value triplet that comprises a property or a value of the first information object.

7. The method of claim 6, wherein the generated object-property-value triplet is incorporated into a Resource Definition Framework (RDF) graph.

8. The method of claim 1, wherein the definition of the instance comprises a name of the instance.

9. The method of claim 1, wherein the definition of the instance comprises an identifier of the concept associated with the instance, wherein the instance structure further comprises the identifier of the concept, and wherein the method further comprises using the identifier of the concept in the instance structure to determine, from the user ontological model, the concept that is associated with the instance.

10. The method of claim 9, wherein the main ontological model comprises the concept.

11. The method of claim 9, further comprising receiving, by the interface from the user, a definition of the concept, wherein the user ontological model further comprises a concept structure based on the definition of the concept.

12. The method of claim 11, wherein the definition of the concept comprises an identifier of a parent concept associated with the concept.

13. The method of claim 12, wherein the main ontological model comprises the parent concept.

14. The method of claim 1, wherein the definition of the instance comprises a natural language text string, wherein the instance structure comprises the natural language text string, and wherein the node corresponds to the natural language text string.

15. The method of claim 1, further comprising identifying, to the user, parts of the text corresponding to generated information objects.

46. The method of claim 9, wherein the user ontological model comprises the concept.

16. A system comprising:

an interface configured to receive, from a user, a definition of an instance of a concept, wherein the instance is specific to a particular field of knowledge of the user; and

a hardware processor configured to;

in response to receipt of the definition, generate a user ontological model based on the definition, wherein the user ontological model is created for the user and the particular field of knowledge of the user, wherein a main ontological model provided to the user does not comprise ontological objects specific to the particular field of knowledge, wherein the user ontological model and the main ontological model are for use in natural language processing of natural language text, wherein the user ontological model comprises an instance structure that is based on the definition, and wherein the instance structure has an identifier;

access a semantic-syntactic tree generated from a text provided by the user, wherein the semantic-syntactic tree comprises a plurality of connected nodes, and wherein at least one node among the plurality of connected nodes comprises a reference to the identifier of the instance structure;

identify the instance structure in the user ontological model based on the reference to the identifier in the semantic-syntactic tree;

in response to identification of the instance structure, generate a first information object for the node in the semantic-syntactic tree based on the instance structure; and

perform natural language processing of the text based on the first information object.

17. The system of claim 16, wherein the hardware processor is further configured to:

identify a fragment in the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node;

determine, from the ontological rule and from the fragment, a property of the instance structure;

determine, from the ontological rule and from the fragment, a value of the property; and

assign the value to a same property of the first information object as the property of the instance structure.

18. The system of claim 16, wherein the hardware processor is further configured to:

identify a fragment of the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node; and

generate a second information object corresponding to the pattern of nodes and to the fragment.

19. The system of claim 18, wherein the hardware processor is further configured to:

determine, from the ontological rule and from the fragment, an ontological object property of an ontological object in the user ontological model or the main ontological model corresponding to the second information object; and

set a value of a property of the second information object to a value of a same property as the first information object.

20. The system of claim 18, wherein the hardware processor is further configured to:

determine, from the ontological rule and from the fragment, a property of the instance structure; and

set a value of a property of the first information object to a value of a same property as the second information object.

21. The system of claim 16, wherein the hardware processor is further configured to generate an object-property-value triplet that comprises a property or a value of the first information object.

22. The system of claim 21, wherein the generated object-property-value triplet is incorporated into a Resource Definition Framework (RDF) graph.

23. The system of claim 16, wherein the definition of the instance comprises a name of the instance.

24. The system of claim 16, wherein the definition of the instance comprises an identifier of the concept associated with the instance, wherein the instance structure further comprises the identifier of the concept, and wherein the hardware processor is further configured to use the identifier of the concept in the instance structure to determine, from the user ontological model, the concept that is associated with the instance.

25. The system of claim 24, wherein the main ontological model comprises the concept.

26. The system of claim 24, wherein the hardware processor is further configured to receive, by the interface from the user, a definition of the concept, and wherein the user ontological model further comprises a concept structure based on the definition of the concept.

27. The system of claim 26, wherein the definition of the concept comprises an identifier of a parent concept associated with the concept.

28. The system of claim 27, wherein the main ontological model comprises the parent concept.

29. The system of claim 16, wherein the definition of the instance comprises a natural language text string, wherein the instance structure comprises the natural language text string, and wherein the node corresponds to the natural language text string.

30. The system of claim 16, wherein the hardware processor is further configured to identify, to the user, parts of the text corresponding to generated information objects.

31. A non-transitory computer-readable storage medium having instructions stored therein that, when executed by a hardware processor, cause the hardware processor to:

receive, by an interface from a user, a definition of an instance of a concept, wherein the instance is specific to a particular field of knowledge of the user;

in response to receipt of the definition, generate, by the hardware processor, a user ontological model based on the definition, wherein the user ontological model is created for the user and the particular field of knowledge of the user, wherein a main ontological model provided to the user does not comprise ontological objects specific to the particular field of knowledge, wherein the user ontological model and the main ontological model are for use in natural language processing of natural language text, wherein the user ontological model comprises an instance structure that is based on the definition, and wherein the instance structure has an identifier;

access a semantic-syntactic tree generated from a text provided by the user, wherein the semantic-syntactic tree comprises a plurality of connected nodes, and wherein at least one node among the plurality of connected nodes comprises a reference to the identifier of the instance structure;

identify the instance structure in the user ontological model based on the reference to the identifier in the semantic-syntactic tree;

in response to identification of the instance structure, generate a first information object for the node in the semantic-syntactic tree based on the instance structure; and

perform, by the hardware processor, natural language processing of the text based on the first information object.

32. The computer readable storage medium of claim 31, wherein the instructions are further to cause the hardware processor to:

identify a fragment in the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node;

determine, from the ontological rule and from the fragment, a property of the instance structure;

determine, from the ontological rule and from the fragment, a value of the property; and

assign the value to a same property of the first information object as the property of the instance structure.

33. The computer readable storage medium of claim 31, wherein the instructions are further to cause the hardware processor to:

identify a fragment of the semantic-syntactic tree, wherein the fragment matches a pattern of nodes defined by an ontological rule in the user ontological model or the main ontological model, and wherein the fragment comprises the node; and

generate a second information object corresponding to the pattern of nodes and to the fragment.

34. The computer readable storage medium of claim 33, wherein the instructions are further to cause the hardware processor to:

determine, from the ontological rule and from the fragment, an ontological object property of an ontological object in the user ontological model or the main ontological model corresponding to the second information object; and

set a value of a property of the second information object to a value of a same property as the first information object.

35. The computer readable storage medium of claim 33, wherein the instructions are further to cause the hardware processor to:

determine, from the ontological rule and from the fragment, a property of the instance structure; and

set a value of a property of the first information object to a value of a same property as the second information object.

36. The computer readable storage medium of claim 31, wherein the instructions are further to cause the hardware processor to generate an object-property-value triplet that comprises a property or a value of the first information object.

37. The computer readable storage medium of claim 36, wherein the generated object-property-value triplet is incorporated into a Resource Definition Framework (RDF) graph.

38. The computer readable storage medium of claim 31, wherein the definition of the instance comprises a name of the instance.

39. The computer readable storage medium of claim 31, wherein the definition of the instance comprises an identifier of the concept associated with the instance, wherein the instance structure further comprises the identifier of the concept, and wherein the instructions are further to cause the hardware processor to use the identifier of the concept in the instance structure to determine, from the user ontological model, the concept that is associated with the instance.

40. The computer readable storage medium of claim 39, wherein the main ontological model comprises the concept.

41. The computer readable storage medium of claim 39, wherein the instructions are further to cause the hardware processor to receive, by the interface from the user, a definition of the concept, and wherein the user ontological model further comprises a concept structure based on the definition of the concept.

42. The computer readable storage medium of claim 41, wherein the definition of the concept comprises an identifier of a parent concept associated with the concept.

43. The computer readable storage medium of claim 42, wherein the main ontological model comprises the parent concept.

44. The computer readable storage medium of claim 31, wherein the definition of the instance comprises a natural language text string, wherein the instance structure comprises the natural language text string, and wherein the node corresponds to the natural language text string.

45. The computer readable storage medium of claim 31, wherein the instructions are further to cause the hardware processor to identify, to the user, parts of the text corresponding to generated information objects.