Auto playlist generation with multiple seed songs

• US 6,987,221 B2
• Filed: 05/30/2002
• Issued: 01/17/2006
• Est. Priority Date: 05/30/2002
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Platt, John C.
• Primary Examiner(s): Fletcher, Marlon T.
• Assistant Examiner(s): Warren, David S.
• Application Number:
  US10/158,674
• Publication Number:
  US 20030221541A1
• Time in Patent Office: 1,328 Days
• Field of Search: 84600-609, 846/15, 846/47, 846/53, 846/34, 846/66, 700/94, 708/172, 707/102, 707/104.1
• US Class Current: 84/601
• CPC Class Codes: G06F 16/4387 by the use of playlists
• 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 16/634 Query by example, e.g. query by humming
• CPC Class Codes: G06F 16/639 using playlists
• 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 16/683 using metadata automatically derived from the content
• CPC Class Codes: G10H 1/0058 Transmission between separate instruments or between individual components of a musical system G10H1/0083 takes precedence
• CPC Class Codes: G10H 2240/061 MP3, i.e. MPEG-1 or MPEG-2 Audio Layer III, lossy audio compression
• CPC Class Codes: G10H 2240/245 ISDN [Integrated Services Digital Network]
• CPC Class Codes: G10H 2240/295 Packet switched network, e.g. token ring circuit-switched networks, e.g. traditional analog telephone transmission G10H2240/241
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

The present invention relates to systems and/or methods that generate playlist(s) for a library or collection of media items via selecting a plurality of seed items, at least one of which is an undesirable seed item. Some of the seed items are desirable indicating that a user prefers additional media items similar to the desirable seed items and others are undesirable indicating that the user prefers additional media items dissimilar to the undesirable seed items. Additionally, the seed items can be weighted to establish a relative importance of the seed items. The invention compares media items in the collection with the seed items and determines which media items are added into the playlist by computation of similarity metrics or values. The playlist can be regenerated by adding desirable seed items to the playlist and removing media items from the playlist (e.g., undesirable seed items).

Claims

1. A system that facilitates generating a playlist comprising:

a media analyzer that receives a plurality of seed media items and obtains metadata corresponding to the plurality of seed media items; and

a playlist generator that receives the metadata obtained from the seed media items, receives at least one candidate media item and its corresponding metadata, computes a similarity value for each of the plurality of seed media items based at least in part on metadata associated with each of the plurality of seed media items and metadata associated with the at least one candidate media item, employs a preference function on the computed similarity values for each of the plurality of seed media items to compute a final similarity value for the at least one candidate media item, and generates a playlist based at least in part on one or more final similarity values.

2. The system of claim 1, the plurality of seed media items comprising at least one undesired seed media item.

3. The system of claim 1, the plurality of seed media items being a media type selected from the group comprising audio, video, electronic books and images.

4. The system of claim 1, the metadata of the plurality of seed media items being a part of the plurality of seed media items.

5. The system of claim 1, the metadata being obtained from analyzing the plurality of seed media items.

6. The system of claim 1, wherein at least one seed media item is a desired seed media item, whereby the playlist is generated to add desirable media items to the playlist based in part on the metadata of the desired seed media item.

7. The system of claim 1, wherein at least one seed media item is an undesired seed media item whereby the playlist is regenerated in part by removing media items from the playlist based in part on metadata associated with the undesired seed media item.

8. The system of claim 1, further employing at least one user-defined seed media item.

9. The system of claim 1, further comprising a media database that maintains the collection of media items and provides metadata corresponding to the collection of media items.

10. The system of claim 9, the playlist generator computing final similarity values for the collection of media items.

11. The system of claim 10, the playlist generator further operative to sort the playlist according to the final similarity values.

12. The system of claim 9, further comprising a media player that plays media items from the collection of media items according to the playlist.

13. A user interface that facilitates playlist generation comprising:

a media library pane that displays information about a media collection and adds and/or removes selected tracks to and from a playlist; and

a playlist pane that displays the playlist, the playlist being regenerated by a playlist generator on each addition or removal of selected tracks and other tracks similar thereto and being a function of the plurality of added tracks and removed tracks;

the playlist generator calculating a final similarity value for one or more candidate tracks by way of using a preference function to aggregate calculated similarity values for each added and removed track, and determining if a candidate track will be included in the playlist based at least in part on the final similarity value.

14. The user interface of claim 13, the playlist pane further displaying removed tracks from the playlist.

15. The user interface of claim 13, the displayed information being organized hierarchically and including artist name, album name and track name.

16. The user interface of claim 13, the media library pane initiates a preview of the selected tracks.

17. The user interface of claim 13, the playlist pane clears tracks from the playlist.

18. The user interface of claim 13, the playlist pane saves the playlist to a file.

19. The user interface of claim 13, further comprising a media player that plays the tacks identified in the playlist.

20. The user interface of claim 13, the playlist pane assigns weights to the selected tacks.

21. A system that facilitates generation of a playlist comprising:

a data store that maintains a collection of one or more media items and metadata corresponding to each media item; and

a playlist generating system that receives a plurality of seed media items and automatically generates and/or regenerates a playlist for the collection according to the metadata associated with added or removed seed media items to effect generation of a new playlist that corresponds to the added or removed seed media items, the playlist generating system determining a final similarity value for one or more candidate media items based at least in part on a preference function that aggregates calculated similarity values for each added and removed seed media item.

22. The system of claim 21, the seed media items comprising at least one undesired seed media item.

23. The system of claim 21, the one or more media items being a media type from the group comprising music, audio books, movies and documents.

24. The system of claim 21, the playlist generating system comprising:

a similarity subsystem that obtains metadata corresponding to the plurality of seed media items, obtains metadata for the one or more media items, and computes final similarity values for each candidate media item of the one or mote media items based on the metadata for that candidate media item and the seed media item metadata; and

a playlist generating subsystem that determines which media items are added to or removed from the playlist.

25. The system of claim 24, the similarity subsystem further comprising a seed item input subsystem that performs preprocessing on the plurality of seed items.

26. The system of claim 21, the plurality of seed media items having associated user-defined weights.

27. The system of claim 21, the plurality of seed media items having associated target preferences.

28. The system of claim 26, the similarity subsystem further individually computes a plurality of similarity values as a function of the media item metadata for the candidate media item and the seed media item metadata, scales the similarity metrics by the associated weights and aggregates the similarity metrics into the final similarity value.

29. A method that facilitates playlist generation, the method comprising:

computing a plurality of similarity values between a plurality of seed media items and a candidate media item;

multiplying each of said plurality of similarity values by associated individual seed media item weighting factors;

aggregating the multiplied similarity values into a final similarity metric;

automatically generating a new playlist by adding or removing candidate media items most similar or dissimilar to the plurality of seed media items, respectively.

30. The system of claim 29, the plurality of seed media items comprising at least one undesirable seed media item.

31. The method of claim 29, wherein computing the plurality of similarity values comprises:

comparing attributes of the candidate media item and attributes of a seed media item;

generating a difference vector based on the comparison; and

computing a similarity value from the difference vector.

32. The method of claim 29, wherein the weighting factors indicate the strength of desirability and/or undesirability for associated seed media items.

33. The method of claim 29, wherein the weighting factors are computed from target preferences.

34. The method of claim 29, further comprising adding or removing the candidate media item to a playlist based on the final similarity value exceeding a threshold amount.

35. The method of claim 34, iteratively proceeding with adding or removing the candidate media item to a playlist based on the final similarity value exceeding a threshold amount for a number of candidate media items of a media collection.

36. A method that facilitates playlist generation comprising:

displaying a collection of media items;

selecting a desirable seed media item from the collection of media;

adding the desirable seed media item to a playlist;

automatically regenerating the playlist based at least in part on metadata associated with the desirable seed media item, the playlist selectively including similar media items from to collection of media items, whereby the similar media items included on the playlist are automatically selected based on similarities to the desirable seed media item;

displaying the playlist;

selecting an undesirable seed item from the playlist;

removing the undesirable seed media item from the playlist; and

automatically regenerating the playlist at least in part by employing a preference function to compute a final similarity value from a plurality of similarity values based on metadata associated with the desirable seed media item and the undesirable seed media item, the playlist selectively including media items similar to the desired seed media item and selectively removing media items similar to the undesired seed media item.

37. The method of claim 36, further comprising adding one or more additional desirable seed media items to the playlist and regenerating the playlist at least in part by employing a preference function based on metadata associated with the one or more additional desirable seed media items.

38. The method of claim 36, further comprising adding one or more additional undesirable seed media items to the playlist and regenerating to playlist at least in part by employing a preference function based on metadata associated with the one or more additional undesirable seed media items.

39. The method of claim 36, further comprising playing media items according to the playlist.

40. A method that facilitates playlist generation comprising:

displaying a collection of media items;

selecting a desirable seed media item from the collection of media items;

adding the desirable seed media item to a playlist;

generating the playlist based at least in part on metadata associated with the desirable seed media item, the playlist selectively including similar media items from the collection of media items;

displaying the playlist;

selecting additional desirable seed media item from the playlist;

adding the additional desirable seed media item to the playlist; and

regenerating the playlist by automatically adding media items that are similar to the additional desirable seed media item based at least in part by utilizing a preference function to compute a final similarity value from a plurality of similarity values based on metadata associated with the desirable seed media item and the additional desired seed media item.

41. The method of claim 40, further comprising adding one or more additional desirable seed media items to the playlist and regenerating the playlist based at least in part on metadata associated with the one or more additional desirable seed media items.

42. The method of claim 40, further comprising adding one or more additional undesirable seed media items to the playlist and regenerating the playlist based at least in part on metadata associated with the one or more additional undesirable seed media items.

43. The method of claim 40, further comprising playing media items according to the playlist.

44. A computer readable medium storing computer executable instructions of components operable to perform a method that facilitates playlist generation, comprising:

a component for computing a plurality of similarity metrics for a plurality of seed media items and a candidate media item to produce a plurality of similarity values;

a component for multiplying the plurality of similarity values by individual seed media item weighting factors;

a component for aggregating the multiplied similarity values into a final similarity value;

a component that automatically generates a new playlist by adding or removing candidate items most similar or dissimilar to the plurality of seed media items, respectively.

45. A data packet comprising:

information associated with a playlist generation system, the information comprising a plurality of final similarity values, respective final similarity values comprising an aggregate of a plurality of seed media item weighted similarity values based at least in part on metadata associated with seed and candidate media items to facilitate automatic regeneration of a new playlist by automatically adding or removing candidate media items based on their respective final similarity values.