Self-adaptive method and system for providing a user-preferred ranking order of object sets

• US 6,370,526 B1
• Filed: 05/18/1999
• Issued: 04/09/2002
• Est. Priority Date: 05/18/1999
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Agrawal, Rakesh, Arning, Andreas, Seiffert, Roland, Srikant, Ramakrishnan
• Primary Examiner(s): Breene, John
• Assistant Examiner(s): RAYYAN, SUSAN F
• Application Number:
  US09/314,479
• Time in Patent Office: 1,057 Days
• Field of Search: 707/5
• US Class Current: 1/1
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A method and system for presenting a group of objects in a ranking order. Objects are ranked according to user preferences by first observing the access order of a related group of objects in relation to a predetermined access hypothesis. A user preference model is then adapted to correspond to any deviations between the access order and the access hypothesis for the related group of objects. Next, object preferences are calculated for each of the objects to be ranked according to the preference model. The group of objects is then presented to the user in an order corresponding to the calculated object preferences. The preference model is adaptively updated, unbeknownst to the user, in the normal course of accessing the presented objects.

Claims

1. A method for presenting a plurality of objects in a ranking order, the method comprising the steps of:

observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

adapting a preference model in response to any deviation between said access order and said access hypothesis, said preference model comprising a preference feature vector, having object features and associated preference feature values, and a plurality of object feature vectors, having object features, for corresponding objects in said second plurality of objects, and associated preference feature values;

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects is related to said first plurality of objects;

presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences; and

thereafter continually adaptively updating the preference model as each plurality of objects are observed and processed.

2. The method according to claim 1, wherein said object preferences are calculated as a function of the preference feature values of said object feature vectors.

3. The method according to claim 2, wherein said object preferences are increased for objects accessed in an order earlier than that predicted by said access hypothesis and for objects having preference features similar to said earlier accessed objects.

4. The method according to claim 3, wherein said object preferences are increased by:

calculating a bonus value for said objects;

then

applying said bonus value to the preference feature values of said object feature vectors.

5. The method according to claim 4, wherein said bonus value relates to the difference between said access order and said access hypothesis.

6. The method according to claim 4, wherein the magnitude of said bonus value depends upon the type of access to said first plurality of objects.

7. The method according to claim 6, wherein said type of access includes at least one of reading, deleting and moving said first plurality of objects.

8. The method according to claim 2, wherein said object preferences are decreased for objects accessed in an order later than that predicted by said access hypothesis and for objects having preference features similar to said later accessed objects.

9. The method according to claim 8, wherein said object preferences are decreased by:

calculating a penalty value for said objects;

then

applying said penalty value to the preference feature values of said object feature vectors.

10. The method according to claim 9, wherein said penalty value relates to the difference between said access order and said access hypothesis.

11. The method according to claim 9, wherein the magnitude of said penalty value depends upon the type of access to said first plurality of objects.

12. The method according to claim 11, wherein said type of access includes at least one of reading, deleting and moving said first plurality of objects.

13. The method according to claim 1, wherein said access hypothesis relates to the relative position of an object within the presentation of said first plurality of objects.

14. The method according to claim 1, wherein said access hypothesis is equal to an offset value, identical for all objects of said first plurality of objects.

15. The method according to claim 1, wherein the calculation of said access hypothesis is dynamically selectable between at least one of:

said access hypothesis being related to the relative position of an object within the presentation of said first plurality of objects; and

said access hypothesis being equal to an offset value, identical for all objects of said first plurality of objects.

16. The method according claim 1, wherein said first and second plurality of objects are electronic mail messages existing within an electronic messaging system.

17. The method according claim 1, wherein said first and second plurality of objects are objects retrieved from a database.

18. The method according to claim 1 wherein the preference model is adaptively updated by dynamically determining object features in the preference feature vector and in the object feature vector.

19. A computer system, comprising:

a housing;

a processor mounted within said housing and processing digital data;

memory mounted within said housing for storing digital data and coupled to said processor;

a display coupled to said processor and said memory to present information derived from digital data processed by said processor;

a system interface coupled to said display and said processor for accessing the information presented on said display; and

a control program stored in said memory that, when executed on said processor, generates a preference model based upon a user'"'"'s access actions on a first group of objects, said objects being presented on said display and accessed through said system interface when executed on said processor, and causes a second group of related objects to be presented in a ranked order on said display based upon said preference model.

20. The system according to claim 19 wherein the control program scaling the object preferences of said first plurality of objects by an aging factor.

21. The system according to claim 20 wherein said aging factor is equal to the value F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of object.

22. A system for presenting a plurality of objects in a ranked order, comprising:

a housing;

a processor mounted within said housing and processing digital data;

memory mounted within said housing for storing digital data and coupled to said processor;

a display coupled to said processor and said memory to present information derived from digital data processed by said processor;

a system interface coupled to said display and said processor for accessing the information presented on said display; and

a control program stored in said memory and accessible by said processor to direct the processing of digital data by said processor,

said control program and said processor cooperating, when said control program is executing on said processor, in observing an access order of a first plurality of objects, presented on said display and accessed through said system interface, in relation to an access hypothesis for those objects,

adapting a preference model, stored in said memory, in response to any deviation between said access order and said access hypothesis,

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects is related to said first plurality of objects,

presenting said second plurality of objects on said display in a ranking order corresponding to said calculated object preferences, and

scaling the object preference of said first plurality of objects by an aging factor.

23. The system according to claim 22 wherein said aging factor is equal to the value of F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of objects.

24. A data processing program for presenting a plurality of objects in a ranked order based upon object preferences, said program executing in a data processing system comprising software code portions for:

observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

scaling the object preference of said first plurality of objects by an aging factor;

adapting a preference model in response to any deviation between said access order and said access hypothesis;

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects is related to said first plurality of objects; and

presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences.

25. The program according to claim 24 wherein said aging factor is equal to the value F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of objects.

26. An article of manufacture comprising:

a computer usable medium, having computer readable code modules embedded therein for presenting a plurality of objects in a ranked order based upon object preference information, said computer readable code modules including

a computer readable code module for observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

a computer readable code module for scaling the object preference of said first plurality of objects by an aging factor;

a computer readable code module for adapting a preference model in response to any deviation between said access order and said access hypothesis;

a computer readable code module for calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects is related to said first plurality of objects; and

a computer readable code module for presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences.

27. The article of manufacture according to claim 26 wherein said aging factor is equal to the value F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of objects.

28. A method for presenting a plurality of objects in a ranking order, the method comprising the steps of:

observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

adapting a preference model in response to any deviation between said access order and said access hypothesis;

scaling the object preferences of said first plurality of objects by an aging factor

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects also includes said first plurality of objects; and

presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences.

29. A method for presenting a plurality of objects in a ranking order, the method comprising the steps of:

observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

adapting a preference model in response to any deviation between said access order and said access hypothesis;

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects includes said first plurality of objects;

presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences; and

scaling the object preferences of said first plurality of objects by an aging factor wherein said aging factor is equal to the value F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of object.

30. A method for presenting a plurality of objects in a ranking order, the method comprising the steps of:

observing an access order of a first plurality of objects in relation to an access hypothesis for those objects;

adapting a preference model in response to any deviation between said access order and said access hypothesis;

calculating an object preference for each of the objects of a second plurality of objects according to said preference model, wherein said second plurality of objects includes said first plurality of objects;

presenting said second plurality of objects in a ranking order corresponding to said calculated object preferences; and

scaling the object preferences of said first plurality of objects by an aging factor wherein said aging factor is equal to the value F^P, where 0<

=F<

=1 and P is equal to the number of times an object preference has been calculated for a particular one of said first plurality of object.