Determining the similarity of music using cultural and acoustic information
There is disclosed a method and storage media for determining the similarity between a first music collection and a second music collection. The similarity may be determined by comparing metadata associated with the first music collection and metadata associated with the second music collection. A plurality of comparison algorithms may be used including at least one of a comparison algorithm based on acoustic metadata and a comparison algorithm based on cultural metadata. The result of the plurality of comparison algorithms may be combined.
1. A method for determining a similarity of a target music collection and a reference music collection with respect to a specific user, comprising:
comparing metadata associated with the target music collection and metadata associated with the reference music collection using a plurality of comparison algorithms including at least one comparison algorithm based on acoustic metadata and at least one comparison algorithm based on cultural metadata, each comparison algorithm producing a result
automatically developing a basis vector from a basis set, the basis set including a plurality of audio tracks representative of the specific user's musical preferences, the basis set being all or a portion of the specific user's music collection, wherein automatically developing the basis vector comprises:
comparing each track in the basis set to each other track in the basis set using each of the plurality of comparison algorithms
averaging the results for each comparison algorithm
scaling the averaged results to form the basis vector
weighting the results of the two or more comparison algorithms using the basis vector
combining the weighted results of the two or more comparison algorithms.
2. The method of claim 1 , further comprising
receiving a request from a requestor to compare the target music collection and the reference music collection
reporting the combined results to the requestor.
3. The method of claim 1 , wherein the comparing, weighting, and combining are done automatically without user intervention.
4. The method of claim 1 , wherein the reference music collection is all or a portion of the specific user's music collection.
5. The method of claim 1 , wherein the at least one comparison algorithm based on acoustic metadata is a plurality of comparison algorithms based on acoustic metadata.
6. The method of claim 1 , wherein the target music collection is one of a first track, a first album, and a first plurality of tracks.
7. A storage medium having instructions stored thereon which, when executed by a processor, will cause the processor to perform actions comprising:
comparing metadata associated with a target music collection and metadata associated with a reference music collection using a plurality of comparison algorithms including at least one comparison algorithm based on acoustic metadata and at least one comparison algorithm based on cultural metadata, each comparison algorithm producing a result
automatically developing a basis vector from a basis set, the basis set including a plurality of audio tracks representative of the specific user's musical preferences, the basis set being all or a portion of the specific user's music collection, wherein automatically developing the basis vector comprises:
comparing each track in the basis set to each other track in the basis set using each of the plurality of comparison algorithms
averaging the results for each comparison algorithm
scaling the averaged results to form the basis vector
weighting the results of the two or more comparison algorithms using the basis vector
combining the weighted results of the two or more comparison algorithms.
8. The storage medium of claim 7 , the actions performed further comprising
receiving a request from a requestor to compare the target music collection and the reference music collection
reporting the combined results to the requestor.
9. The storage medium of claim 7 , wherein the comparing, weighting, and combining are done automatically without user intervention.
10. The storage medium of claim 7 , wherein the reference music collection is all or a portion of the specific user's music collection.
11. The storage medium of claim 7 , wherein the at least one comparison algorithm based on acoustic metadata is a plurality of comparison algorithms based on acoustic metadata.
12. The storage medium of claim 7 , wherein the target music collection is one of a first track, a first album, and a first plurality of tracks.
13. A computing device for comparing music, comprising:
a processor
memory coupled to the processor
a storage device storing instructions that, when executed by the processor, cause the computing device to perform actions comprising:
receiving a request from a requestor, the request identifying a target music collection, a reference music collection, and a specific user
comparing metadata associated with the target music collection and metadata associated with the reference music collection using a plurality of comparison algorithms including at least one comparison algorithm based on acoustic metadata and at least one comparison algorithm based on cultural metadata, each comparison algorithm producing a result
deriving a basis vector specific to the user from a basis set including audio tracks representative of the specific user's musical preferences, the basis set being all or a portion of the specific user's music collection, wherein deriving the basis vector comprises:
comparing each track in the basis set to each other track in the basis set using each of the plurality of comparison algorithms
averaging the results for each comparison algorithm
scaling the averaged results to form the basis vector
weighting the results of the plurality of comparison algorithms using the basis vector
combining the weighted results of the plurality of comparison algorithms
providing the combined weighted result to the requestor.
14. The computing device of claim 13 , wherein the target music collection is one of a track, an album, and a plurality of tracks.
15. The computing device claim 14 , wherein the reference music collection is all or a part of the specific user's music collection.
16. A method of recommending music, comprising:
a web server sending a request to a similarity server, the request identifying a target music collection, a reference music collection, and a specific user
the similarity server determining the similarity between the target music collection and the reference music collection with respect to the specific user, wherein determining further comprises:
comparing metadata associated with the target music collection and metadata associated with the reference music collection using a plurality of comparison algorithms including at least one comparison algorithm based on acoustic metadata and at least one comparison algorithm based on cultural metadata, each comparison algorithm producing a result
developing a basis vector from a basis set, the basis set including a plurality of audio tracks representative of the specific user's musical preferences, the basis set being all or a portion of the specific user's music collection, wherein automatically developing the basis vector comprises:
comparing each track in the basis set to each other track in the basis set using each of the plurality of comparison algorithms
averaging the results for each comparison algorithm
scaling the averaged results to form the basis vector
weighting the results of the plurality of comparison algorithms using the basis vector
combining the weighted results of the plurality of comparison algorithms to produce a similarity result
the similarity server providing the similarity result to the web server
the web server determining to recommend, or not recommend, the target music collection to the specific user based on the similarity result.