Method and apparatus for utilizing user feedback to improve signifier mapping

• US 8,255,541 B2
• Filed: 10/30/2008
• Issued: 08/28/2012
• Est. Priority Date: 05/23/2000
• Status: Expired due to Term

• Current Assignee: OpenTV, Inc. (Kudelski SA)
• Original Assignee: RPX Corporation
• Inventors: Reisman, Richard
• Primary Examiner(s): Nguyen, Quang N.
• Application Number:
  US12/261,419
• Publication Number:
  US 20090119286A1
• Time in Patent Office: 1,398 Days
• Field of Search: 709/203, 709/204, 709/217, 709/227, 709/245, 707/3, 707/5
• US Class Current: 709/227
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results

Abstract

An apparatus for finding resources on a network comprises a finder server having access to: (a) a database including: (i) an index of resources available on network of interconnected computers on which a plurality of resources reside; and (ii) information regarding user feedback gathered from previous operations of the apparatus by a user and plural previous users; and (b) a learning system operable to access and learn from information contained on the database. The finder server is operable to locate, in response to entry by the user of a resource identity signifier, a single intended target resource intended by the user to uniquely correspond to the resource identity signifier, from among a plurality of resources located on the network, by: receiving a resource identity signifier from the user; accessing the database to determine, based on the information in the database, which, if any, of the indexed resources is likely to be the intended target resource; and directing a computer of the user so as to cause that computer to connect the user to the address of the resource, if any, determined as likely to be the intended target resource.

Claims

1. A method, comprising:

processing a user input comprising a signifier for a target resource with reference to a heuristic knowledge base utilizing a processor of a computing platform to determine a possible target resource, wherein the possible target resource is determined at least in part without regard to any comparison of linguistic relationships of the signifier to content and/or metadata contained within the possible target resource, wherein the heuristic knowledge base is dynamically updated utilizing the processor in view of feedback relating to one or more prior responses to one or more corresponding signifier inputs from one or more users, and wherein the signifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention; and

updating the heuristic knowledge base in accordance with one or more express inputs received from one or more additional users identifying additional and/or desired pairings of one or more target resources to be associated with one or more respective signifiers.

2. The method of claim 1, wherein said processing the user input comprises determining that the possible target resource comprises the target resource at least in part in response to deriving from the heuristic knowledge base information indicative of a predetermined confidence level associated with the user input.

3. The method of claim 2, further comprising:

updating the heuristic knowledge base with feedback from multiple users relating to the accuracy of said determining that the possible target resource comprises the target resource.

4. The method of claim 1, wherein said target resource comprises a command.

5. The method of claim 1, wherein said target resource comprises an internet resource.

6. The method of claim 1, further comprising executing a defined action corresponding to the possible target resource utilizing the processor of the computing platform at least in part in response to the possible target resource having a highest confidence level value of a plurality of possible target resources.

7. The method of claim 1, wherein the target resource comprises a command and wherein a defined action comprises an execution of the command.

8. The method of claim 1, wherein the target resource comprises an internet resource and wherein a defined action comprises a presentation of the internet resource.

9. The method of claim 1, wherein said processing the user input comprises presenting a list of one or more possible target resources ranked with respect to the user input at least in part in response to none of the possible target resources having an associated confidence level equal to or greater than a pre-selected level.

10. The method of claim 9, further comprising executing a defined action corresponding to a highest confidence level value with respect to the user input for a possible target resource.

11. The method of claim 1, further comprising provisionally updating the heuristic knowledge base with a possible new user input signifier to target resource relationship.

12. The method of claim 1, wherein said processing the user input comprises displaying advertising content at least in part during said determining the possible target resource.

13. The method of claim 12, wherein the advertising content comprises interstitial advertising.

14. The method of claim 1, wherein said processing the user input comprises determining that the possible target resource comprises the target resource at least in part in response to deriving from the heuristic knowledge base a predetermined confidence level associated with the user input.

15. A method for finding a target, comprising:

processing a plurality of user inputs comprising one or more signifiers for a target resource from a respective plurality of users with reference to a heuristic knowledge base utilizing a processor of a computing platform to determine a possible target resource, wherein the possible target resource is determined at least in part without regard to any comparison of linguistic relationships of the one or more signifiers to content and/or metadata contained within the possible target resource, wherein the one or more signifiers comprise natural language expressions processed without regard to any linguistic constraints imposed by a standard and/or convention;

updating said heuristic knowledge base at least in part in response to receiving feedback from the plurality of users; and

updating said heuristic knowledge base in accordance with one or more express inputs received from one or more of the plurality of users identifying additional and/or desired pairings of one or more target resources to be associated with one or more respective signifiers.

16. The method of claim 15, wherein said processing the plurality of user inputs comprises determining that the possible target resource comprises the target resource at least in part in response to deriving from the heuristic knowledge base information indicative of a confidence level associated with the possible target resource.

17. The method of claim 15, further comprising:

updating the heuristic knowledge base with feedback from the plurality of users relating to the accuracy of said determining that the possible target resource comprises the target resource.

18. The method of claim 15, further comprising:

determining which of a plurality of stored possible target resources is likely to comprise the target resource based at least in part on deriving from the heuristic knowledge base information indicative of a confidence level relating to one or more previous usages of the one or more signifiers for the target resource and results of one or more attempted connections to possible target resources;

updating the heuristic knowledge base to add information usable to increase a confidence level associated with a mapping between an individual target resource of the one or more additional target resources and an associated signifier at least in part in accordance with an individual input from the one or more express inputs; and

updating the heuristic knowledge base to add information usable to decrease the confidence level associated with a mapping between another individual target resource of the one or more additional target resources and an associated signifier at least in part in accordance with another individual input from the one or more express inputs.

19. The method of claim 15, wherein said target resource comprises a command.

20. The method of claim 15, wherein said target resource comprises an internet resource.

21. The method of claim 15, wherein said processing the plurality of user inputs comprises presenting a list of one or more possible target resources ranked with respect to the plurality of user inputs at least in part in response to none of the possible target resources having an associated confidence level equal to or greater than a pre-selected level.

22. The method of claim 21, further comprising executing a defined action corresponding to an individual possible target resource having a highest confidence level value among the one or more possible target resources.

23. The method of claim 15, wherein the target resource comprises a command and wherein a defined action associated with the target resource comprises an execution of the command.

24. The method of claim 15, wherein the target resource comprises an internet resource and a wherein a defined action associated with the target resource comprises a presentation of the internet resource.

25. The method of claim 15, further comprising provisionally updating the heuristic knowledge base with social usage information for a possible new user input/target resource pair relationship.

26. The method of claim 15, wherein said processing the plurality of user inputs comprises displaying advertising content at least in part during said determining the possible target resource.

27. The method of claim 26, wherein the advertising content comprises interstitial advertising.

28. The method of claim 15, wherein said processing the plurality of user inputs comprises determining that the possible target resource comprises the target resource at least in part in response to deriving from the heuristic knowledge base a confidence level associated with the possible target resource.

29. A method of finding an intended target resource, comprising:

receiving a resource identity signifier intended to uniquely correspond to the target resource located on a network, wherein the resource identity signifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention;

accessing database information that includes an index of available resources on the network, and includes information indicative of associated confidence levels and associated social usage of the resource identity signifier from multi-user feedback, both the associated confidence levels and social usage relating to previous results of attempts to connect to intended target resources;

determining which of a plurality of matched indexed resources is likely to be the intended target resource that uniquely corresponds to the resource identity signifier using the associated confidence level of the resource identity signifier, wherein said determining is performed at least in part without regard to any comparison of linguistic relationships of the resource identity signifier to content and/or metadata contained within the plurality of matched indexed resources;

updating the database information to add information usable to increase the confidence level associated with a mapping between the resource identity signifier and an address of the determined intended target resource if feedback indicates that the connected intended target resource corresponds to the resource identity signifier;

updating the database information so as add information usable to decrease the confidence level associated with a mapping between the resource identity signifier and the address of the determined intended target resource if feedback indicates that the connected intended target resource fails to correspond to the resource identity signifier; and

updating the database information provisionally to add an available resource after locating the resource.

30. The method of claim 29, further comprising updating the database to provisionally modify the social usage information for a possible new signifier-to-resource relationship using an add request.

31. The method of claim 29, further comprising:

connecting to a URL corresponding to an indexed resource having a highest confidence level.

32. The method of claim 29, further comprising:

presenting the user with a list of one or more links to possible resources, the list being ordered according to a confidence level, with a resource having a highest confidence level being ranked highest.

33. The method of claim 29, wherein said target resource comprises a command.

34. The method of claim 29, wherein said target resource comprises an internet resource.

35. The method of claim 29, further comprising presenting a list of one or more possible intended target resources, the list being ordered according to the associated confidence level, with a possible intended target resource having a higher confidence level being ranked higher than a possible intended target resource with a lower confidence level, but only if none of the possible intended target resources has an associated confidence level equal to at least a predetermined level.

36. The method of claim 29, wherein the intended target resource is delivered along with advertising.

37. The method of claim 36, wherein the advertising comprises interstitial advertising.

38. The method of claim 29, wherein the information indicative of associated confidence levels comprises one or more confidence levels associated with one or more respective resources.

39. A method of determining, in response to entry of an item identity signifier, an intended target item, comprising:

receiving the item identity signifier;

accessing a database on a finder server on a network;

learning a social usage of the item identity signifier from multi-user feedback inputs based on previous results by the finder server; and

determining an indexed item as likely uniquely corresponding to the item identity signifier based on the social usage of the item identity signifier, wherein the indexed item is determined at least in part without regard to any comparison of linguistic relationships of the item identity signifier to content and/or metadata contained within the indexed item, wherein the item identity signifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention.

40. The method of claim 39, wherein said item comprises a command.

41. The method of claim 39, wherein said item comprises an internet resource.

42. The method of claim 39, further comprising:

presenting a list of one or more possible items, the list being ordered according to an associated confidence level, with an item having a higher confidence level being ranked higher than an item with a lower confidence level, but only if none of the indexed items has an associated confidence level equal to at least a predetermined level.

43. The method of claim 39, further comprising:

executing a defined action corresponding to an indexed item having a highest confidence level value utilizing a processor of a computing platform.

44. The method of claim 39, wherein the item comprises a command and a defined action comprises an action called for by that command.

45. The method of claim 39, wherein the item comprises an internet resource and the defined action comprises the presentation of that internet resource.

46. The method of claim 39, further comprising:

causing the database on the finder server to provisionally update the social usage information for a possible new signifier-to-item relationship.

47. The method of claim 39, further comprising conveying advertising content related to the item.

48. The method of claim 47, wherein the advertising content includes interstitial advertising.

49. A method of interpreting user-inputted identifiers as unique item identifications, comprising:

enabling a population of users to submit identifier-item pairs, an identifier and an item in each pair being associated with one another;

learning a social usage of identifiers associated with unique items by the population of users based at least in part on feedback from previous operations of the method from multiple members of the population; and

for any user in the population of users, mapping an identifier submitted by the user to a single corresponding item based on the social usage of the entered identifier, wherein the identifier is mapped at least in part without regard to any comparison of linguistic relationships of the identifier to content and/or metadata contained within the single corresponding item, wherein the identifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention.

50. The method of claim 49, further comprising generating a list displayed on a computer of one or more links to possible items, the list being ordered according to confidence level, with an item having a highest confidence level being ranked highest.

51. The method of claim 49, further comprising causing a heuristic knowledge base to provisionally update the social usage information for a possible new identifier item pair.

52. The method of claim 49, further comprising presenting a list of one or more possible items, the list being ordered according to an associated confidence level, with a possible item having a higher confidence level being ranked higher than a possible item with a lower confidence level, but only if none of the possible items has an associated confidence level equal to at least a predetermined level.

53. The method of claim 49, further comprising generating a list of one or more links to possible items, the list being ordered according to confidence level, with a possible item having a highest confidence level being ranked highest.

54. The method of claim 49, further comprising updating the social usage provisionally with a possible new identifier-item pair relationship.

55. The method of claim 49, wherein feedback information gathered from a plurality of previous users of the system is used to change an identifier-item pair.

56. The method of claim 49, wherein the feedback information is gathered using clickstream data.

57. The method of claim 49, further comprising determining which of a plurality of stored items likely is the single item based at least in part on information in a knowledge base indicative of a confidence level relating to previous usage and results of attempted connections to intended items;

updating the knowledge base to add information usable to increase the confidence level associated with a mapping between the identifier item pair if a feedback response indicates that a connected item corresponds to the single item; and

updating the knowledge base to add information usable to decrease the confidence level associated with a mapping between the identifier item pair if a feedback response indicates that a connected item does not corresponds to the single item.

58. The method of claim 49, further comprising displaying a related advertisement on a computing platform display.

59. The method of claim 58, wherein the advertisement comprises interstitial advertising.

60. A method of translating user-inputted identifiers into unique item identifications, comprising:

enabling a population of users to submit identifier-item pairs to be associated with one another;

learning a social usage of identifiers associated with unique items by a population of multiple users sharing a defined subject or interest; and

for any user, mapping a user-entered identifier to a single corresponding item known to the user based on the social usage of the entered identifier, wherein the user-entered identifier is mapped at least in part without regard to any comparison of linguistic relationships of the user-entered identifier to content and/or metadata contained within the single corresponding item, wherein the identifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention.

61. The method of claim 60, further comprising generating a list of one or more links to items, the list being ordered according to confidence level, with an item having a highest confidence level being ranked highest.

62. The method of claim 60, further comprising updating a knowledge base provisionally with social usage information for a new identifier-to-item relationship.

63. The method of claim 60, further comprising presenting a list of one or more possible items, the list being ordered according to an associated confidence level, with a possible item having a higher confidence level being ranked higher than a possible item with a lower confidence level, but only if none of the possible items has an associated confidence level equal to at least a predetermined level.

64. The method of claim 60, further comprising updating the social usage provisionally.

65. The method of claim 60, wherein feedback information gathered from a plurality of previous users of the system is used to modify an identifier-to-item association.

66. The method of claim 60, wherein the feedback information is gathered using clickstream data.

67. The method of claim 60, further comprising:

determining which of a plurality of stored items likely is the single item based at least in part on information in a knowledge base indicative of a confidence level relating to previous usage and results of attempted connections to intended items;

updating the knowledge base to add information usable to increase the confidence level associated with a mapping between the identifier item pair if a feedback response indicates that a connected item corresponds to the single item; and

updating the knowledge base to add information usable to decrease the confidence level associated with a mapping between the identifier item pair if a feedback response indicates that a connected item does not corresponds to the single item.

68. The method of claim 60, wherein the unique item comprises an associated advertisement.

69. The method of claim 68, wherein the advertisement comprises interstitial advertising.

70. A method of connecting to an intended target resource, comprising:

receiving at a server a resource identity signifier intended to uniquely correspond to a resource located on a network;

accessing database information on the server that includes an index of available resources on the network, and includes information indicative of associated confidence levels and associated social usage of the resource identity signifier derived from multi-user feedback, both the associated confidence levels and social usage relating to previous results of attempts to connect to intended target resources;

determining which of a plurality of matched indexed resources is likely to be the intended target resource that uniquely corresponds to the recognized resource identity signifier using the associated confidence level associated with a mapping between the resource identity signifier and an address of the matched target resource, wherein the intended target resource is determined at least in part without regard to any comparison of linguistic relationships of the resource identity signifier to content and/or metadata contained within the intended target resource, wherein the confidence is for the signifier-item association, wherein the resource identity signifier comprises a natural language expression processed without regard to any linguistic constraints imposed by a standard and/or convention;

updating the database information using a multi-user feedback to add information usable to increase the confidence level associated with a mapping between the resource identity signifier and an address of the determined intended target resource if feedback indicates that the connected intended target resource corresponds to the resource identity signifier;

updating the database information using a multi-user feedback to add information usable to decrease the confidence level associated with a mapping between the resource identity signifier and the address of the determined intended target resource if feedback indicates that the connected intended target resource fails to correspond to the resource identity signifier;

updating the database information provisionally to add an available resource after locating the resource;

updating the database information to modify the social usage using the feedback;

generating a list displayed on a computer of one or more links to a possible matched resource, the list being ordered according to confidence level, with a resource having a highest confidence level being ranked highest; and

providing a selective option to enable any user to update the database with an available resource.

71. The method of claim 70, wherein said determining which of a plurality of matched indexed resources is likely to be the intended target resource comprises displaying at least one associated advertisement on a display of a computing platform.

72. The method of claim 71, wherein the advertisement comprises interstitial advertising.