Dynamic identification of a new set of media items responsive to an input mediaset

• US 7,693,887 B2
• Filed: 02/01/2005
• Issued: 04/06/2010
• Est. Priority Date: 02/01/2005
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Strands Incorporated (CRIBIS Holding SRL)
• Inventors: McLaughlin, Matthew R.
• Primary Examiner(s): Alam; Shahid A
• Assistant Examiner(s): Bromell; Alexandria Y
• Application Number:
  US11/048,950
• Publication Number:
  US 20060173910A1
• Time in Patent Office: 1,890 Days
• Field of Search: 707/3, 707/104.1, 707/616, 707/913
• US Class Current: 707/616
• CPC Class Codes: G06F 16/4387 by the use of playlists
• CPC Class Codes: Y10S 707/913 Multimedia

Abstract

Systems and methods are disclosed for creating a list of media items that are similar to an input mediaset comprising at lease one input media item, optionally together with at least one input parameter that constrains the output in scope or number of items. Media items include music, videos, pictures and other such items that can be associated or grouped together by a human into a list or mediaset. The mediaset is formed by a human applying virtually any objective and/or subjective criteria. Those criteria need not be made explicit; a relation among the selected items is implicit in the list per se. Alternatively, objective criteria such as metadata can be used.

Claims

1. A method for dynamically generating a set of media items related to an input set of media items comprising the steps of:

(a) receiving an input mediaset wherein the input mediaset comprises a list of unique identifiers, each unique identifier corresponding to a respective media item

(b) comparing the input mediaset list of unique media item identifiers to a database of predetermined mediasets, each of the predetermined mediasets comprising respective lists of unique media item identifiers;

wherein said comparing step comprises identifying which of the predetermined mediasets includes at least one unique media item identifier that matches one of the unique media item identifiers within the input mediaset list of unique media item identifiers; and

wherein said matching is with regard to the unique media item identifiers and without regard to a descriptor, characteristic, classification or attribute of any of the media items themselves;

(c) merging all the identified predetermined mediasets into a single merged list of similar media items; and

(d) creating an output set of media items by selecting one or more media items from the merged list of similar media items and;

(e) delivering a list of the output set of media items.

2. The method in claim 1, wherein the items in the input mediaset were selected by one or more persons so as to form the input mediaset.

3. The method in claim 1, wherein the input mediaset have been at least partially ordered as deemed appropriate by a human.

4. The method in claim 1, wherein said creating the output set of media items includes selecting a randomized subset of the top items of the merged list.

5. The method in claim 1, wherein the database of mediasets resides on a network server, and the merging and creating an output set occurs on that network server.

6. The method in claim 1, wherein the database of mediasets resides on or coupled to a network server, and the merging and selection operations occur on a client computer operated by a user.

7. The method in claim 1, and further comprising storing the input mediaset in the said database.

8. The method in claim 1, wherein the media items include music songs and tracks.

9. The method in claim 1, wherein the media items include video items.

10. The method in claim 1, wherein such media items include images.

11. A storage medium bearing computer executable instructions for carrying out the method of claim 1.

12. A method according to claim 1 including computing a discrimination weight for each media item listed in the input mediaset.

13. A method according to claim 1 wherein said merging all the identified predetermined mediasets into a single ranked list of similar media items includes:

for each of the identified predetermined mediasets, computing a metric of its similarity to the input mediaset.

14. A method according to claim 13 wherein said computing a metric of similarity to the input mediaset comprises:

computing a discrimination weight for each media item in the input mediaset; and

for each of the identified predetermined mediasets, computing the metric of its similarity to the input mediaset as a function of the discrimination weights of the media items co-occurring in the input mediaset and the predetermined mediaset.

15. A method according to claim 13 wherein said merging all the identified predetermined mediasets into a single ranked list of media items includes:

assembling a list of all media items that occur in the identified predetermined mediasets but do not occur in the input mediaset; and

computing an item rank weight for each item on the assembled list.

16. A method according to claim 15 wherein said computing an item rank weight comprises:

determining in which of the retrieved mediasets the item occurs; and

calculating the item rank weight as a function of the similarity metrics of the retrieved mediasets in which the item occurs to the input mediaset.

17. The method in claim 1, further comprising receiving one or more input parameters that are used to restrict the output set of media.

18. The method in claim 17, wherein at least one parameter is specified by the user.

19. The method in claim 1, wherein said creating an output set includes filtering the items of the ranked list according to a predefined parameter.

20. The method in claim 19 wherein the predefined parameter is specified by the user.

21. The method in claim 19, wherein the predefined parameter includes one of an artist, album, genre, or sub-genre.

22. A kiosk for distributing digital music items comprising:

means for receiving from a user an input list of one or more music items selected by the user;

means for converting the input list to a list of unique identifiers, each unique identifier corresponding to a respective music item on the input list;

means for accessing a database comprising a plurality of previously stored lists of music items;

wherein each of the stored lists comprises a list of unique identifiers, each unique identifier corresponding to a respective music item;

computer means for identifying at least one unique identifier among the unique identifiers listed in the previously stored lists of music items that does not correspond to a music item on the input list and has a predefined relationship to the input list of music items;

wherein the predefined relationship is determined by first identifying which of the previously stored lists of music items includes at least one unique identifier that matches one of the unique identifiers corresponding to the input list of music items;

wherein said matching is determined with regard to the unique identifiers and without regard to a descriptor, characteristic, classification or attribute of any of the corresponding media items themselves; and

then including within the defined relationship all of the unique descriptors listed in the identified previously stored lists of music items;

means for merging all the identified previously stored lists of music items into a single merged list of similar media items; and

output means for recommending at least one music item to the user wherein the recommended music item corresponds to the identified unique identifier.

23. A kiosk for distributing digital music according to claim 22 wherein the database is located remote from the kiosk and is accessed via a network.

24. A kiosk for distributing digital music according to claim 22 wherein the computer means is located remote from the kiosk and is accessed via a network.