DISCOVERING AND SCORING RELATIONSHIPS EXTRACTED FROM HUMAN GENERATED LISTS

• US 20080250039A1
• Filed: 04/03/2008
• Published: 10/09/2008
• Est. Priority Date: 04/04/2007
• Status: Active Grant

• Current Assignee: Intertrust Technologies Corporation (Fidelio Acquisition Co LLC)
• Original Assignee: Seeqpod Incorporated
• Inventors: Franks, Kasian, Podowski, Raf, Muldoon, Mike
• Granted Patent: US 8,108,417 B2
• Time in Patent Office: Days
• US Class Current: 707/100
• CPC Class Codes: G06F 16/3331 Query processing
• CPC Class Codes: G06F 16/3334 Selection or weighting of terms from queries, including natural language queries
• CPC Class Codes: G06F 16/4387 by the use of playlists
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

A computer-implemented system and method for extracting Human Generated Lists from an electronic database is described. The system searches for objects of the same class within a context window to identify Human Generated Lists and stores them to an archive. The archive may be used to generate a relationship network. The system generates variable length data vectors to represent the relationships between the objects within each Human Generated List. This relationship network can then be queried to discover relationships between the objects in the Human Generated Lists and to provide related objects as recommendations.

Claims

1. A computer-implemented method of generating an electronic collection of human generated lists, the method comprising:

obtaining an electronic database;

selecting a first record in the electronic database; and

determining whether the first record contains at least one human generated list, wherein if the record contains at least one human generated list,

storing data corresponding to the at least one human generated list to an electronic collection of human generated lists.

2. The computer-implemented method of claim 1, wherein the method further comprises repeating the steps of selecting, determining, and storing the data.

3. The computer-implemented method of claim 1, wherein determining whether first record contains at least one human-generated list comprises:

defining a context window over a at least some of the electronic database; and

identifying two or more objects as being in a same class within the defined context window.

4. The computer-implemented method of claim 3, wherein the class comprises a plurality of objects of a same type.

5. The computer-implemented method of claim 4, wherein the class comprises a musical album, and the objects comprise at least one of audio data, textual data, and image data associated with the musical album.

6. The computer-implemented method of claim 1, wherein the human generated list comprises data indicative of a human-compiled collection of non-randomly ordered objects.

7. The computer-implemented method of claim 1, wherein the electronic database comprises a plurality of web pages and the first portion of the electronic database comprises one of the plurality of web pages.

8. The computer-implemented method of claim 1, wherein the step of determining whether the first portion of the electronic database contains at least one human generated list comprises determining whether more than one object from a class is present in the first portion of the electronic database.

9. The computer-implemented method of claim 8, wherein the class comprises a list of at least one of:

songs, musical albums, musical artists, musical genres, movies, videos, actors, actresses, directors, movie genres, books, authors, video games, and video game genres.

10. The computer-implemented method of claim 1, wherein the first human generated list is a ranked list.

11. The computer-implemented method of claim 1, wherein the electronic database contains a plurality of portions, and wherein at least one of the portions does not contain a human generated list.

12. A computer-implemented method of providing a recommendation in response to a user query, the method comprising:

providing a human generated list archive comprising a plurality of ordered human generated lists, wherein each human generated list comprises a plurality of objects;

receiving a user query;

matching the query against the human generated list archive to retrieve a recommendation; and

providing the recommendation to the user.

13. The computer-implemented method of claim 12, further comprising providing a list archive comprising non-human generated lists, wherein matching the query further comprises matching the query against the non-human generated list archive.

14. The computer-implemented method of claim 12, wherein the plurality of object comprises data indicative of a human-compiled collection of non-randomly ordered objects.

15. The computer-implemented method of claim 12, wherein the query object comprises data related to at least one of a song title, an artist, an album title, or a genre.

16. The computer-implemented method of claim 12, wherein constructing relationship vectors comprises analyzing a distance value between objects in the HGL.

17. The computer-implemented method of claim 16, wherein the distance value is based at least in part on a number of words separating the objects in the HGL.

18. The computer-implemented method of claim 17, wherein the number of words excludes stop-words.

19. The computer-implemented method of claim 12, further comprising:

determining a first relationship between objects in a first human generated list of the plurality of ordered human generated lists;

determining a second relationship between objects in a second human generated list of the plurality of ordered human generated lists; and

determining a third relationship between objects in the first and second human generated lists based upon the first and second relationships,

wherein providing the recommendation comprises returning the at least one recommended object from the human generated lists based at least in part on the first, second, and third relationships.

20. The computer-implemented method of claim 19, wherein determining the first relationship comprises selecting a core object and generating a relationship vector related to the core object.

21. The computer-implemented method of claim 20, wherein the relationship vector is based on distance values computed with respect to an associated object.

22. The computer-implemented method of claim 21, wherein the distance value is based at least in part on a number of words separating the core object and the associated object.

23. A computer-implemented system for providing recommended media objects in response to a query object comprising:

a storage containing a human generated list network, the human generated list network comprising vector representations of direct relationships between media objects;

a vector processor configured to determine a plurality of indirect relationships between the media objects based on the vector representations of direct relationships;

a score generator configured to generate a plurality of scores, the scores associated with the direct and indirect relationships between media objects;

an object selector configured to select a set of the media objects based upon the scores associated with the direct and indirect relationships between the media objects and the query object; and

a display configured to display the selected set of media objects.

24. A computer-implemented system for providing recommended media objects in response to a user query, comprising:

means for providing a human generated list archive comprising a plurality of ordered human generated lists, wherein each human generated list comprises a plurality of objects;

means for receiving a user query;

means for matching the query against the human generated list archive to retrieve a recommendation; and

means for providing the recommendation to the user.

25. The computer-implemented system of claim 24, wherein the means for providing a human generated list archive comprises at least one of a relationship processor, a relationship database, an information database, or a web server.

26. The computer-implemented system of claim 25, wherein the means for receiving the user query comprises a web server, or wherein the means for matching comprises a network generator, or wherein the means for providing the recommendation comprises one or more of a query results processor and web server.