Station library creaton for a media service
View Patent ↗A machine may form all or part of a network-based system configured to provide media service to one or more user devices. The machine may be configured to define a station library within a larger collection of media files. In particular, the machine may access metadata that describes a seed that forms the basis on which the station library is to be defined. The machine may determine a genre composition for the station library based on the metadata. The machine may generate a list of media files from the metadata based on a relevance of each media file to the station library. The machine may determine the relevance of each media file based on a similarity of the media file to the genre composition of the station library as well as a comparison of metadata describing the media file to the accessed metadata that describes the seed.
1. A method comprising:
accessing seed metadata, the seed metadata describing a seed, the seed being a basis on which a station library is to be defined;
generating a station descriptor profile based on the seed metadata, the station descriptor profile defining a genre composition of the station library;
generating a candidate set based on the seed metadata, the candidate set comprising a plurality of candidate media files;
for each candidate media file in the candidate set:
computing a similarity score associated with the candidate media file, the similarity score including a measure of similarity between the candidate media file and the station descriptor profile;
computing one or more boost values associated with the candidate media file based on candidate metadata describing the candidate media file; and
computing a relevancy score associated with the candidate media file based on the similarity score and the one or more boost values associated with the candidate media file, the relevancy score providing a basis for selecting the candidate media file for inclusion in the station library; and
machine-generating a station set including a portion of the candidate set selected based on the relevancy scores associated with the plurality of candidate media files included in the candidate set, the machine-generated station set defining the station library by referencing each candidate media file in the portion of the candidate set,
wherein the station descriptor profile includes one or more focus genre profiles and each candidate media file includes a file genre profile, the one or more focus genre profiles and the file genre profile each including respective multiple genre values and a weight assigned to each genre value, each genre value corresponding to a genre, each weight indicating a percentage of the corresponding genre value relative to the other genre values in the respective multiple genre values, and
computing the similarity score associated with the candidate media file includes:
computing one or more focus-level similarity scores by comparing, for each focus genre profile in the station descriptor profile: (a) the respective multiple genre values and the corresponding weights of the focus genre profile, and (b) the respective multiple genre values and the corresponding weights of the file genre profile of the candidate media file; and
selecting the highest focus-level similarity score to be the similarity score associated with the candidate media file.
2. The method of claim 1 , wherein:
the generating of the station descriptor profile includes:
extracting a seed genre profile from the seed metadata, the genre profile describing one or more seed genres; and
deriving the one or more focus genre profiles from the seed genre profile.
3. The method of claim 1 , wherein:
a dedicated portion of the station set is allocated for each focus genre profile in the station descriptor profile; and
the machine-generating of the station set includes populating each dedicated portion with candidate media files based on the relevancy scores associated with each candidate media file in the candidate set.
4. The method of claim 1 , wherein:
the generating of the candidate set includes identifying and selecting candidate media files having an associative relationship with the seed.
5. The method of claim 1 , wherein:
the computing of the one or more boost values includes computing a language boost value associated with the candidate media file based on a comparison of a first language attribute and a second language attribute, the first language attribute corresponding to a language of the candidate media file, the second language attribute corresponding to a language of the seed.
6. The method of claim 1 , wherein:
the computing of the one or more boost values includes computing a related artist boost value associated with the candidate media file based on a relationship type and weight of an associative relationship between artists of the seed and candidate media file.
7. The method of claim 1 , wherein:
the computing of the one or more boost values includes computing a recency artist boost value associated with the candidate media file based on a temporal value attribute associated with the seed in combination with a temporal attribute of the candidate media file included in candidate metadata describing the candidate media file.
8. The method of claim 1 , wherein:
the computing of the one or more boost values includes computing a co-occurrence boost value based on one or more co-occurrences of the seed and a candidate.
9. The method of claim 1 , wherein:
the computing of the one or more boost values includes computing a usage boost value based on an analysis of user activity with respect to the candidate media file.
10. The method of claim 1 further comprising:
determining a regional popularity value associated with each candidate media file in the candidate set, the regional popularity value including a measure of popularity of the candidate media file in a geographical region associated with the seed.
11. The method of claim 10 , wherein:
the determining of the relevancy score associated with the candidate media file includes aggregating the similarity score, the one or more boost values, and the regional popularity value.
12. The method of claim 1 , further comprising:
ranking the plurality of candidate media files in the candidate set according to the relevancy score; and
selecting the portion of the candidate set based on the ranking of the plurality of candidate media files.
13. The method of claim 1 , further comprising:
creating a seed artist set, the seed artist set including a plurality of media files corresponding to an artist associated with the seed; and
incorporating the seed artist set into the station set.
14. The method of claim 1 , further comprising:
excluding a candidate media file from the candidate set based on an attribute-level similarity score of its primary genre not transgressing a threshold similarity score.
15. The method of claim 1 , further comprising:
excluding a candidate media file from the candidate set based on an attribute level similarity score of its primary mood not transgressing a threshold similarity score.
16. The method of claim 1 , further comprising:
excluding a candidate media file from the candidate set based on the similarity score associated with the candidate media file not transgressing a threshold similarity score value.
17. The method of claim 1 further comprising:
excluding a media file from the station set based on application of one or more constraints to the station set.
18. A non-transitory machine-readable storage medium comprising instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:
accessing seed metadata, the seed metadata describing a seed, the seed being a basis on which a station library is to be defined;
generating a station descriptor profile based on the seed metadata, the station descriptor profile defining a genre composition of the station library;
generating a candidate set based on the seed metadata, the candidate set comprising a plurality of candidate media files;
for each candidate media file in the candidate set:
computing a similarity score associated with the candidate media file, the similarity score including a measure of similarity between the candidate media file and the station descriptor profile;
computing one or more boost values associated with the candidate media file based on candidate metadata describing the candidate media file; and
computing a relevancy score associated with the candidate media file based on the similarity score and the one or more boost values associated with the candidate media file, the relevancy score providing a basis for selecting the candidate media file for inclusion in the station library; and
machine-generating a station set including a portion of the candidate set selected based on the relevancy scores associated with the plurality of candidate media files included in the candidate set, the machine-generated station set defining the station library by referencing each candidate media file in the portion of the candidate set,
wherein the station descriptor profile includes one or more focus genre profiles and each candidate media file includes a file genre profile, the one or more focus genre profiles and the file genre profile each including respective multiple genre values and a weight assigned to each genre value, each genre value corresponding to a genre, each weight indicating a percentage of the corresponding genre value relative to the other genre values in the respective multiple genre values, and
computing the similarity score associated with the candidate media file includes:
computing one or more focus-level similarity scores by comparing, for each focus genre profile: (a) the respective multiple genre values and the corresponding weights of the focus genre profile, and (b) the respective multiple genre values and the corresponding weights of the file genre profile of the candidate media file; and
selecting the highest focus-level similarity score to be the similarity score associated with the candidate media file.
19. A system comprising:
one or more processors of a machine; and
a machine-readable medium storing instructions that, when executed by the one or more processors of a machine, cause the machine to perform operations comprising:
accessing seed metadata, the seed metadata describing a seed, the seed being a basis on which a station library is to be defined;
generating a station descriptor profile based on the seed metadata, the station descriptor profile defining a genre composition of the station library;
generating a candidate set based on the seed metadata, the candidate set comprising a plurality of candidate media files;
for each candidate media file in the candidate set:
computing a similarity score associated with the candidate media file, the similarity score including a measure of similarity between the candidate media file and the station descriptor profile;
computing one or more boost values associated with the candidate media file based on candidate metadata describing the candidate media file; and
computing a relevancy score associated with the candidate media file based on the similarity score and the one or more boost values associated with the candidate media file, the relevancy score providing a basis for selecting the candidate media file for inclusion in the station library; and
machine-generating a station set including a portion of the candidate set selected based on the relevancy scores associated with the plurality of candidate media files included in the candidate set, the machine-generated station set defining the station library by referencing each candidate media file in the portion of the candidate set,
wherein the station descriptor profile includes one or more focus genre profiles and each candidate media file includes a file genre profile, the one or more focus genre profiles and the file genre profile each including respective multiple genre values and a weight assigned to each genre value, each genre value corresponding to a genre, each weight indicating a percentage of the corresponding genre value relative to the other genre values in the respective multiple genre values, and
computing the similarity score associated with the candidate media file includes:
computing one or more focus-level similarity scores by comparing, for each focus genre profile: (a) the respective multiple genre values and the corresponding weights of the focus genre profile, and (b) the respective multiple genre values and the corresponding weights of the file genre profile of the candidate media file; and
selecting the highest focus-level similarity score to be the similarity score associated with the candidate media file.