System, method and application of ontology driven inferencing-based personalization systems

• US 20020173971A1
• Filed: 03/28/2001
• Published: 11/21/2002
• Est. Priority Date: 03/28/2001
• Status: Abandoned Application

• Current Assignee: Reuters Limited (The Woodbridge Co. Ltd.)
• Original Assignee: Reuters Limited (The Woodbridge Co. Ltd.)
• Inventors: Stirpe, Paul Alan, Antico, Michael, Pinfold, William John, Slavin, Tim
• Application Number:
  US09/818,953
• Publication Number:
  US 20020173971A1
• Time in Patent Office: Days
• US Class Current: 705/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0255 based on user history

Abstract

The present invention provides a system, method, and applications for providing personalized user experiences based on the use of a core ontology and inferencing over the ontology using rules provided by a domain expert. The population of users may be known to a commerce or information service from external and internal user data sources. Information (data) about this population is brought into a knowledge warehouse designed for on-line analytic processing, and potentially data marts. Data can be sourced from external databases in batch or streaming mode and enhanced with real-time click stream events from internal observed user interactions. A reference ontology is either loaded into the system or defined via a domain expert. The ontology forms the central reference point for data enrichment and precise personalization. Characteristic data is tagged in accordance with direct reference to the nodes of the ontology and may be enhanced via inferencing techniques. This results in enriched and more precise data tagging and equates to discovery of interest domains not directly observed in the initial source data. Definitions of communities can be embedded in the reference ontology thereby allowing the rapid assignment of individuals to collaborative filters or discovered via statistical means using the enriched attributes. Discovery can be fed back into the ontology to add extensions to the ontology. The same reference ontology is used to tag content, which results in a consistent tagging discipline for data and content centered on the reference ontology. Using inference techniques based on the ontology, content may be enriched to discover attributes not explicitly announced in the content descriptions. The enriched data may be mapped to the enriched content resulting in a deeply personalized user experience.

Claims

1. A system for providing personalized content to a user, comprising:

a data warehouse that stores user data corresponding to a user;

an ontology;

an inferencing engine that generates consequences based on information in said data warehouse, wherein said user data is tagged in accordance with said ontology.

2. The system of claim 1, wherein the data warehouse contains healthcare data.

3. The system of claim 1, wherein the data warehouse contains human resource data.

4. The system of claim 1, wherein the data warehouse contains financial data.

5. The system of claim 1, further comprising:

a content store,

wherein content information from said content store is tagged in accordance with said ontology.

6. The system of claim 1, wherein said inferencing engine generates and outputs a personal interest graph (PIG) created for the user based on data rules.

7. The system of claim 5, wherein said inferencing engine generates and outputs a personal interest graph (PIG) created for the user based on data rules,

said system further comprising:

a display for displaying selective information from said content store based at least in part on the PIG.

8. The system of claim 1, wherein the inferencing engine generates and outputs a list of weighted nodes.

9. The system of claim 5, said display providing a personalized view of said content for said user.

10. The system of claim 5, said display providing a personalized view of said content regarding said user for a third party.

11. The system of claim 1, wherein said user data includes click stream data.

12. The system of claim 1, wherein said user data includes source data.

13. The system of claim 1, wherein said user data includes explicit data.

14. The system of claim 1, wherein said user data includes implicit data.

15. The system of claim 1, further comprising a third party user obtaining a personalized view of said user, wherein the third party user is displayed information relating at least in part to said user'"'"'s personalized view.

16. The system of claim 15, wherein the third party provides information to said user related to said displayed information.

17. The system of claim 15, wherein the third party provides information regarding said user to another, other than said user.

18. The system of claim 1, further comprising:

a data mart that receives tagged user data and an analytics console that analyzes said tagged user data in at least one of said data mart and said data warehouse.

20. The method of claim 19, wherein said drawing conclusions step is performed by at least one inferencing engine.

21. The method of claim 19, wherein said receiving user data step includes receiving healthcare data related to said user.

22. The method of claim 19, wherein said receiving user data step includes receiving human resource data related to said user.

23. The method of claim 19, wherein said receiving user data step includes receiving financial data related to said user.

24. The method of claim 19, further comprising the step of:

generating a personal interest graph (PIG) regarding a user based on data rules.

25. The method of claim 19, further comprising the steps of:

generating and outputting a list of weighted nodes.

26. The method of claim 19, further comprising the step of:

displaying said conclusions to said user.

27. The method of claim 19, further comprising the step of:

displaying said conclusions to a third party.

28. The method of claim 19, further comprising the steps of:

receiving content;

tagging said content in accordance with said ontology.

29. The method of claim 19, further comprising the step of:

enhancing said user data with at least one of click stream data, source data, explicit data, and implicit data.

30. The method of claim 19, further comprising the steps of:

separately storing said tagged user data in a data mart, and

analyzing said separately stored tagged user data.

44. The system according to claim 1, wherein said user is de-identified in said data warehouse.

45. The method according to claim 19, said receiving step further comprising the steps of:

receiving user data relating to a de-identified user; and

,

authenticating said de-identified user.

48. The system of claim 47, wherein said consequences are a weighted list.

49. The system of claim 47, wherein said consequences are a content information graph.

50. The system according to claim 47, further comprising:

a data warehouse that stores tagged user data; and

a second inferencing engine that generates consequences based on said tagged user data.

51. The system according to claim 50, further comprising:

a comparator that compares the consequences of from said first inferencing engine with the consequences from said second inferencing engine.

19. A method for drawing conclusions for personalized content relating to a user, comprising the steps of:

receiving user data corresponding to a user;

tagging said user data in accordance with an ontology; and

drawing conclusions over at least said tagged user data.

31. A system for drawing conclusions for personalized content relating to a user, comprising:

means for receiving user data corresponding to a user;

means for tagging said user data in accordance with an ontology; and

means for drawing conclusions over at least said tagged user data.

32. The system of claim 31, wherein said means for drawing conclusions further comprises:

means for drawing inferences.

33. The system of claim 31, further comprising:

means for generating a personal interest graph (PIG) regarding a user based on data rules.

34. The system of claim 31, further comprising:

means for generating outputting a list of weighted nodes.

35. The system of claim 31, further comprising:

means for displaying said conclusions to said user.

36. The system of claim 31, further comprising:

means for displaying said conclusions to a third party.

37. The system of claim 31, further comprising:

means for receiving content;

means for tagging said content in accordance with said ontology.

38. The system of claim 31, further comprising:

means for enhancing said user data with at least one of click stream data, source data, explicit data, and implicit data.

39. The method of claim 31, further comprising:

means for separately storing said tagged user data in a data mart, and

means for analyzing said separately stored tagged user data.

46. The system according to claim 31, further comprising:

means for receiving user data relating to a de-identified user; and

,

means for authenticating said de-identified user.

40. A computer-readable medium for storing a program, said program for drawing conclusions for personalized content relating to a user, said program having the steps of:

receiving user data corresponding to a user;

tagging said user data in accordance with an ontology; and

drawing conclusions over at least said tagged user data.

41. A computer-readable medium for storing a data structure, said data structure comprising:

a first portion storing user data tagged in accordance with an ontology;

a second portion storing a weighting value associated with said user data.

42. The computer-readable medium according to claim 41, said second portion being part of a list of weighted nodes.

43. The computer-readable medium according to claim 41, said data structure forming a personalized interest graph.

47. A system for providing tagged content comprising:

a content store that stores content information;

an ontology;

a first inferencing engine that generates consequences based on information in said content store, wherein said content information is tagged in accordance with said ontology.

52. A method for drawing conclusions for content comprising the steps of:

receiving content information;

tagging said content information in accordance with an ontology; and

drawing first conclusions over at least said tagged content information.

53. The method according to claim 52, further comprising the steps of:

storing tagged user data in a data warehouse; and

drawing second conclusions over at least said tagged user data.

54. The method according to claim 53, further comprising the step of:

comparing the consequences of from said drawing first conclusions step with the consequences of said second conclusions step.

55. The method according to claim 52, wherein said first conclusions are a weighted list.

56. The method according to claim 52, wherein said first conclusions are a content information graph.

58. The system according to claim 57, further comprising:

means for storing tagged user data in a data warehouse; and

means for drawing second conclusions over at least said tagged user data.

59. The system according to claim 58, further comprising:

means for comparing the consequences of from said means for drawing said first conclusions with the consequences of said means for drawing said second conclusions.

57. A system for drawing conclusions for content comprising:

means for receiving content information;

means for tagging said content information in accordance with an ontology; and

means for drawing first conclusions over at least said tagged content information.