Method for content recommendation

• US 7,958,119 B2
• Filed: 02/04/2008
• Issued: 06/07/2011
• Est. Priority Date: 03/31/2007
• Status: Active Grant

• Current Assignee: Sony Deutschland GmbH (Sony Group Corp.)
• Original Assignee: Sony Deutschland GmbH (Sony Group Corp.)
• Inventors: Feduszczak, Tomasz, Kemp, Thomas, Hagg, Wilhelm, Eggink, Jana, Zimmer, Taufan
• Primary Examiner(s): BOCCIO, VINCENT F
• Application Number:
  US12/025,425
• Publication Number:
  US 20090013002A1
• Time in Patent Office: 1,219 Days
• Field of Search: 707/999.002, 707/999.104, 707/732, 707/733, 707/748, 715/700, 715/716, 715/727, 715/744, 715/745, 715/746, 715/747
• US Class Current: 707/732
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: G06F 16/337 Profile generation, learning or modification
• CPC Class Codes: G06F 16/40 of multimedia data, e.g. slideshows comprising image and additional audio data retrieval of still image data G06F16/50; retrieval of audio data G06F16/60; retrieval of video data G06F16/70
• CPC Class Codes: G06F 16/48 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/637 Administration of user profiles, e.g. generation, initialization, adaptation or distribution
• CPC Class Codes: G06F 16/68 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 3/04817 using icons graphical or visual programming using iconic symbols G06F8/34
• CPC Class Codes: G06F 3/04842 Selection of displayed objects or displayed text elements G06F3/0482 takes precedence
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G11B 27/105 of operating discs
• CPC Class Codes: G11B 27/34 Indicating arrangements indicating means incorporated in magazine or cassette G11B23/046 and G11B23/0875; indicating measured values in general G01D

Abstract

A method for content recommendation for a user, wherein a song or a video is recommended to the user and a confidence measure is determined for the recommended song. The confidence measure is displayed to the user, so the user may get more confidence into the recommendation of the system.

Claims

1. A method, implemented by an information processing apparatus, for content recommendation, comprising:

determining, by the information processing apparatus, a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item;

determining, by the information processing apparatus, a confidence measure for said content recommendation, said confidence measure being descriptive of an estimated quality of the determined content recommendation;

communicating said confidence measure or a derivative thereof; and

requesting feedback for at least a part of said content recommendation, wherein the feedback indicates a liking or disliking of said content recommendation and includes different options that change based on the confidence measure.

2. The method according to claim 1, wherein

said feedback relates to meta data attributes of said content recommendation, and

a level of detail of the meta data attributes depends on the confidence measure.

3. The method according to claim 1, wherein, when the confidence measure is above a threshold and the feedback indicates a disliking of said content recommendation, the feedback includes feedback relating to meta data attributes.

4. The method according to claim 1, wherein, when the confidence measure is below a threshold, the feedback only includes feedback indicating a liking or disliking of said content recommendation and does not include meta data attributes.

5. The method according to claim 2, wherein, when the confidence measure is below a threshold and the feedback indicates a disliking of said content recommendation, the feedback includes feedback relating to said meta data attributes, and, when the confidence measure is below a threshold and the feedback indicates a liking of said content recommendation, the feedback includes feedback relating to said meta data attributes.

6. The method according to claim 1, further comprising:

receiving feedback for at least a part of said content recommendation;

inquiring for further feedback; and

adapting said profile based on said feedback and/or said further feedback.

7. The method according to claim 6, wherein, when said further feedback indicates a dislike of said content recommendation and the feedback indicates a general liking of said content recommendation but not at the moment, the profile is not adapted.

8. The method according to claim 1, wherein, said determining of said confidence measure is based on an estimated quality of the profile.

9. The method according to claim 8, wherein said estimated quality is determined depending on an amount of interactions.

10. The method according to claim 8, wherein said estimated quality is determined depending on coherence of the profile.

11. The method according to claim 1, further comprising:

determining a virtual best matching content item based on said profile, said virtual best matching content item matching all meta data attributes given feedback for as close as possible; and

calculating a distance measure between a content item of said content recommendation and said virtual best matching content item, wherein said confidence measure is determined based on said distance measure.

12. The method according to claim 1, wherein a green light is lighted or a green feedback button is displayed, when said confidence measure has a high value, and

a red light is lighted or a red feedback button is displayed, when said confidence measure has a low value.

13. The method according to claim 1, wherein said confidence measure is used to explain why a certain content recommendation was selected.

14. The method according to claim 1, wherein the feedback is indirect feedback.

15. A device for content recommendation, comprising:

a data processor configured to determine a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item, and

to determine a confidence measure for said content recommendation, wherein said confidence measure is descriptive of an estimated quality of the determined content recommendation;

a display configured to display said confidence measure or a derivative thereof; and

a feedback part configured

to input feedback for at least a part of said content recommendation,

the feedback indicating a liking or disliking of said content recommendation, and

to input further feedback, the further feedback including different options that change based on said confidence measure,

wherein said data processor updates said profile based on said feedback and/or said further feedback.

16. A non-transitory computer-readable medium storing computer readable instructions thereon that when executed by a content recommendation device cause the content recommendation device to perform a method, comprising:

displaying a graphical user interface that includes

a selectable negative feedback button associated with a first changeable text field, wherein said first changeable text field indicates that normal negative feedback can be input for a currently played content item,

wherein, in response to a normal negative feedback being input, a confidence measure for the currently played content item being above a threshold, and said feedback indicating a disliking of said currently played content item,

the first changeable text field changes and indicates that a further refined feedback can be input for the currently played content item.

17. The non-transitory computer-readable medium according to claim 16, wherein the graphical user interface further includes a selectable negative feedback button which upon selection causes a negative feedback to be output to a data processor.

18. A non-transitory computer-readable medium storing computer readable instructions thereon that when executed by a content recommendation device cause the content recommendation device to perform a method comprising:

displaying a graphical user interface that includes

at least one positive feedback button associated with a first text field, said positive feedback button allowing input of positive feedback for a currently played song; and

at least one negative feedback button associated with a second text field, said negative feedback button allowing input of negative feedback for said currently played song, wherein an appearance of said first and second text field is based on a confidence measure for said currently played song, said confidence measure being descriptive of an estimated quality of the currently played song.

19. A non-transitory computer-readable medium storing computer readable instructions thereon that when executed by a content recommendation device cause the content recommendation device to perform a method comprising:

determining a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item;

determining a confidence measure for said content recommendation, said confidence measure being descriptive of an estimated quality of the determined content recommendation;

communicating said confidence measure or a derivative thereof; and

requesting for feedback for at least a part of said content recommendation, wherein the feedback indicates a liking or disliking of said content recommendation and includes different options that change based on the confidence measure.

20. A device for content recommendation, comprising:

means for determining a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item;

means for determining a confidence measure for said content recommendation, wherein said confidence measure is descriptive of an estimated quality of the determined content recommendation;

means for communicating said confidence measure or a derivative thereof; and

means for requesting feedback for at least part of said content recommendation, wherein the feedback indicates a liking or disliking of said content recommendation and includes different options that change based on the confidence measure.

21. A method, implemented by an information processing apparatus, for content recommendation, comprising:

determining, by the information processing apparatus, a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item;

determining, by the information processing apparatus, a confidence measure for said content recommendation, said confidence measure being descriptive of an amount of data about a user'"'"'s preferences;

receiving feedback for at least a part of said content recommendation;

requesting for further feedback, wherein the feedback indicates a liking or disliking of said content recommendation and includes different options that change based on the confidence measure.

22. The method according to claim 21, further comprising:

adapting said profile based on said feedback and/or said further feedback.

23. A method, implemented by an information processing apparatus, for content recommendation, comprising:

determining, by the information processing apparatus, a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item;

determining, by the information processing apparatus, a confidence measure for said content recommendation, said confidence measure being descriptive of an estimated quality of the determined content recommendation;

communicating said confidence measure or a derivative thereof; and

requesting feedback for at least a part of said content recommendation, wherein a type of the feedback depends on the confidence measure, wherein said confidence measure or said derivative thereof is communicated simultaneously with requesting feedback for at least part of said content recommendation.

24. The method according to claim 23, wherein

a green light is lighted or a green feedback button is displayed, when said confidence measure has a high value, and

a red light is lighted or a red feedback button is displayed, when said confidence measure has a low value.

25. A device for content recommendation, comprising:

a data processor configured

to determine a content recommendation from a data base of content items based on a profile, wherein said content recommendation includes at least one content item, and

to determine a confidence measure for said content recommendation, wherein said confidence measure is descriptive of an estimated quality of the determined content recommendation;

a graphical user interface configured to display said confidence measure or a derivative thereof; and

at least one positive feedback button associated with a first text field, said at least one positive feedback button allowing input of positive feedback for a currently played song; and

at least one negative feedback button associated with a second text field, said at least one negative feedback button allowing input of negative feedback for said currently played song,

wherein

an appearance of said first and/or second text field depends on a confidence measure for said currently played song, said confidence measure being descriptive of an estimated quality of the currently played song,

said graphical user interface displays said confidence measure or said derivative thereof simultaneously with said at least one positive and said at least one negative feedback button.

26. The device according to claim 25, wherein

an icon associated with a positive result or feeling is displayed when the confidence measure of the currently recommended at least one content item is high, and

an icon associated with a negative result or feeling is displayed when the confidence measure of the currently recommended at least one content item is low.

27. The device according to claim 25, wherein the graphical user interface further includes a graphical item configured

to emit green light when said confidence measure is high, and

to emit red light when said confidence measure is low.

28. The device according to claim 27, wherein said graphical item is configured

to emit green light when said confidence measure is above a first threshold,

to emit red light when said confidence measure is below a second threshold, and

to emit yellow light when said confidence measure lies in between said first and second threshold.