Methods and devices for customizing knowledge representation systems

• US 9,235,806 B2
• Filed: 03/15/2013
• Issued: 01/12/2016
• Est. Priority Date: 06/22/2010
• Status: Active Grant

• Current Assignee: Primal Fusion Inc.
• Original Assignee: Primal Fusion Inc.
• Inventors: Sweeney, Peter Joseph, Ilyas, Ihab Francis
• Primary Examiner(s): Holmes, Michael B
• Application Number:
  US13/844,009
• Publication Number:
  US 20130246328A1
• Time in Patent Office: 1,033 Days
• Field of Search: 706/50
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks

Abstract

Techniques for customizing knowledge representation systems including identifying, based on a plurality of concepts in a knowledge representation (KR), a group of one or more concepts relevant to user context information, and providing the identified group of one more concepts to a user. The KR may include a combination of modules. The modules may include a kernel and a customized module customized for the user. The kernel may accessible via a second KR.

Claims

1. A method comprising:

obtaining user context information associated with a user, wherein the user context information includes information regarding an attribute of the user, information regarding an activity of the user, and/or information provided by the user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR.

2. The method of claim 1, wherein the plurality of concepts in the first KR is a first plurality of concepts, wherein the identified group of one or more concepts is a first group of one or more concepts relevant to the user context information, wherein the customized module is a first customized module, and wherein the method further comprises:

identifying, based on a second plurality of concepts in the second KR, a second group of one or more concepts relevant to the user context information;

providing the identified second group of one or more concepts to the user,

wherein the second KR includes a combination of the kernel and a second customized module, the second customized module being customized for the user.

3. The method of claim 2, wherein there is at least one concept and/or relationship that is included in only one of the first customized module or the second customized module.

4. The method of claim 2, wherein the first customized module includes a first concept and/or first relationship relevant to first activity of the user, and the second customized module includes a second concept and/or second relationship relevant to second activity of the user.

5. The method of claim 2, wherein the first customized module is limited to activity of the user meeting a set of one or more criteria selected from the group consisting of:

activity of the user during specified hours, activity of the user on specified days of the week, activity of the user on specified dates, activity carried out by the user with one or more specified computing devices, activity carried out by the user from one or more specified internet addresses, activity carried out by the user while logged into one or more specified user accounts, and activity of the user within one or more specified geographical regions.

6. The method of claim 5, wherein the set of one or more criteria is a first set, wherein the second customized module is limited to activity of the user failing to meet the first set of one or more criteria, and/or activity of the user meeting a second set of one or more criteria, wherein the second set of one or more criteria differs from the first set and is selected from the group consisting of:

activity of the user during specified hours, activity of the user on specified days of the week, activity of the user on specified dates, activity carried out by the user with one or more specified computing devices, activity carried out by the user from one or more specified internet addresses, activity carried out by the user while logged into one or more user accounts, and activity of the user within one or more geographical regions.

12. The method of claim 1, wherein the kernel includes first and second reference modules, wherein the first reference module is specific to a first information domain and includes concepts relevant to the first information domain, and wherein the second reference module is specific to a second information domain and includes concepts relevant to the second information domain.

13. The method of claim 12, wherein:

the plurality of concepts based on which the identified group of one or more concepts are identified includes one or more concepts of the first reference module and does not include at least one concept of the second reference module that is not included in the first reference module.

14. The method of claim 13, wherein the plurality of concepts in the first KR is a first plurality of concepts, wherein the identified group of one or more concepts is a first group of one or more concepts relevant to the user context information, wherein the content information is first content information, and wherein the method further comprises:

identifying, based on a second plurality of concepts in the first KR, a second group of one or more concepts relevant to the user context information;

identifying second content information corresponding to the identified second group of one or more concepts; and

providing the identified second content information to the user.

15. The method of claim 12, wherein the plurality of concepts based on which the identified group of one or more concepts are identified includes at least one concept of the first reference module and at least one concept of the second reference module that is not included in the first reference module.

16. The method of claim 12, wherein there is a difference between the first and second reference modules.

19. The method of claim 1, wherein providing the identified content information to the user comprises ranking portions of the identified content information based on relevance of the portions of the identified content information to the identified group of one or more concepts.

20. The method of claim 1, wherein the customized module is a first customized module, wherein the first KR includes a second customized module customized for the user, wherein the method further comprises constructing a third customized module customized for the user based on analysis of the first and second customized modules.

21. The method of claim 1, further comprising performing an analysis of the customized module, wherein performing the analysis of the customized module comprises altering the customized module, and wherein the analysis of the customized module is performed at a specified time, periodically, or in response to one or more triggers selected from the group consisting of:

creation of the customized module, modification of the customized module, a number of concepts in the customized module exceeding a threshold number of concepts, and a number of relationships in the customized module exceeding a threshold number of relationships.

27. The method of claim 1, wherein:

the data structure representing the first KR is a graph;

the graph includes a first node, a second node, and an edge between the first and second nodes;

the first node corresponds to a first concept of the first KR;

the second node corresponds to a second concept of the first KR; and

the edge corresponds to a relationship between the first and second concepts and/or represents a relevance of the first concept to the second concept.

7. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the customized module is a first customized module, wherein the modules further include a second customized module, the second customized module being customized for the user, and wherein there is a difference between the first and second customized modules.

8. The method of claim 7, wherein the difference between the first and second customized modules comprises at least one concept and/or relationship that is included in only one of the first customized module or the second customized module.

9. The method of claim 7, wherein the first customized module includes a first concept and/or first relationship relevant to first activity of the user, and wherein the second customized module includes a second concept and/or second relationship relevant to second activity of the user.

10. The method of claim 9, wherein the first customized module is limited to activity of the user meeting a set of one or more criteria selected from the group consisting of:

activity of the user during specified hours, activity of the user on specified days of the week, activity of the user on specified dates, activity carried out by the user with one or more specified computing devices, activity carried out by the user from one or more specified internet addresses, activity carried out by the user while logged into one or more specified user accounts, and/or activity of the user within one or more specified geographical regions.

11. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the method further comprises;

identifying content information corresponding to the identified group of one or more concepts; and

providing the identified content information to the user.

17. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein;

an interest network of the user includes the user context information, the identified group of one or more concepts, and/or one or more additional groups of one or more concepts identified in response to the user context information, and

the method further comprises updating the first KR based on analysis of at least a portion of the interest network.

18. The method of claim 17, wherein updating the first KR based on the analysis of at least the portion of the interest network is performed at a specified time, periodically, or in response to one or more triggers selected from the group consisting of:

the content information being provided, an amount of the provided content information exceeding a threshold amount, the identified group of one or more concepts being provided, and a size of the identified group of relevant concepts exceeding a threshold size.

22. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between a user concept included in the user model and a kernel concept included in the kernel.

23. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between two kernel concepts included in the kernel.

24. The method of claim 23, wherein the kernel does not include any information identifying the relationship between the two kernel concepts included in the kernel.

25. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the user model includes a user concept not included in the kernel.

26. The method of claim 25, wherein:

the user concept is a first user concept,

the user model includes a second user concept not included in the kernel, and

the user model includes a relationship between the first user concept and the second user concept.

28. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the first KR includes elemental concepts and complex concepts, and wherein the complex concepts are obtained by applying synthesis processing to one or more of the elemental concepts.

29. A method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using at least one processor and a data structure representing the first KR,

wherein the plurality of concepts in the first KR include a plurality of elemental concepts, and wherein identifying, based on the plurality of concepts, the group of one or more concepts comprises identifying, from among the plurality of elemental concepts, the group of one or more concepts.

30. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user, wherein the user context information includes information regarding an attribute of the user, information regarding an activity of the user, and/or information provided by the user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information;

providing the identified group of one or more concepts to the user,

wherein the first KR includes a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and providing are performed at least in part by using a data structure representing the first KR.

31. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user, wherein the user context information includes information regarding an attribute of the user, information regarding an activity of the user, and/or information provided by the user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information;

identifying content information corresponding to the identified group of one or more concepts; and

providing the identified content information to the user,

wherein the first KR includes a kernel and a customized module, the kernel accessible via by a second KR, the customized module being customized for the user, and

wherein the identifying and providing are performed at least in part by using a data structure representing the first KR.

32. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein;

an interest network of the user includes the user context information, the identified group of one or more concepts, and/or one or more additional groups of one or more concepts identified in response to the user context information, and

the method further comprises updating the first KR based on analysis of at least a portion of the interest network.

33. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information;

providing the identified group of one or more concepts to the user;

identifying content information corresponding to the identified group of one or more concepts; and

providing the identified content information to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR.

34. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the customized module is a first customized module, wherein the modules further include a second customized module, the second customized module being customized for the user, and wherein there is a difference between the first and second customized modules.

35. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between a user concept included in the user model and a kernel concept included in the kernel.

36. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between two kernel concepts included in the kernel.

37. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes a user concept not included in the kernel.

38. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the first KR includes elemental concepts and complex concepts, and wherein the complex concepts are obtained by applying synthesis processing to one or more of the elemental concepts.

39. A system comprising:

at least one processor configured to perform;

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the plurality of concepts in the first KR include a plurality of elemental concepts, and wherein identifying, based on the plurality of concepts, the group of one or more concepts comprises identifying, from among the plurality of elemental concepts, the group of one or more concepts.

40. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein;

an interest network of the user includes the user context information, the identified group of one or more concepts, and/or one or more additional groups of one or more concepts identified in response to the user context information, and

the method further comprises updating the first KR based on analysis of at least a portion of the interest network.

41. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the method further comprises;

identifying content information corresponding to the identified group of one or more concepts; and

providing the identified content information to the user.

42. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the customized module is a first customized module, wherein the modules further include a second customized module, the second customized module being customized for the user, and wherein there is a difference between the first and second customized modules.

43. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between a user concept included in the user model and a kernel concept included in the kernel.

44. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes information identifying a relationship between two kernel concepts included in the kernel.

45. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the user model includes a user concept not included in the kernel.

46. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the first KR includes elemental concepts and complex concepts, and wherein the complex concepts are obtained by applying synthesis processing to one or more of the elemental concepts.

47. At least one non-transitory computer-readable storage medium storing processor executable instructions that, when executed by at least one processor, cause the at least one processor to perform a method comprising:

obtaining user context information associated with a user;

identifying, based on a plurality of concepts in a first knowledge representation (KR), a group of one or more concepts relevant to the user context information; and

providing the identified group of one or more concepts to the user,

wherein the first KR includes a combination of modules, the modules including a kernel and a customized module, the kernel being accessible via a second KR, the customized module being customized for the user, and

wherein the identifying and the providing are performed at least in part by using the at least one processor and a data structure representing the first KR,

wherein the plurality of concepts in the first KR include a plurality of elemental concepts, and wherein identifying, based on the plurality of concepts, the group of one or more concepts comprises identifying, from among the plurality of elemental concepts, the group of one or more concepts.