Methods and apparatus for determining a mood profile associated with media data
View Patent ↗In an embodiment, a method is provided for determining a mood profile of media data. In this method, mood is determined across multiple elements of mood for the media data to create a mood profile associated with the media data. In some embodiments, the mood profile is then used to determine congruencies between one or more pieces of media data.
1. A method comprising:
accessing, from a database, a first mood vector that describes first media data and specifies a first mood-value pair, the first mood-value pair assigning a first value to a first mood;
accessing, from the database, a second mood vector that describes a second media data and specifies a second mood-value pair, the second mood-value pair assigning a second value to a second mood;
retrieving a first score that quantifies congruency between the first and second moods, the first score being retrieved from a data structure that correlates the first and second moods;
using a processor, generating a second score that quantifies congruency between the first and second mood vectors, the generating the second score being based on the accessed first and second values and the retrieved first score; and
based on the generated second score, adding the first media data to a playlist.
2. The method of claim 1 , wherein:
the retrieved first score is a comparative value that represents an extent to which the first and second moods are similar to each other.
3. The method of claim 1 , wherein:
the data structure that correlates the first and second moods is a mood correlates table in which a cell that indicates a relationship between the first mood and the second mood contains the first score that quantifies the congruency between the first and second moods; and
the retrieving of the first score from the data structure retrieves the first score from the mood correlates table.
4. The method of claim 1 , wherein:
the first score that quantifies the congruency between the first and second moods is generated by a person;
the method further comprising populating a cell of the data structure with the first score in response to the first score being received from the person; and
wherein the retrieving of the first score is from the populated cell of the data structure.
5. The method of claim 1 , wherein:
the first score that quantifies the congruency between the first and second moods is an average of multiple values submitted by multiple users;
the method further comprises populating a cell of the data structure with an average of the multiple values in response to the multiple values being received from the multiple users; and
wherein the retrieving of the first score retrieves the average from the populated cell of the data structure.
6. The method of claim 1 , wherein:
the generated second score is a similarity score that represents extents to which the first and second mood vectors are similar to each other and to which the first and second media data are similar to each other.
7. The method of claim 1 , wherein:
the generating of the second score includes normalizing the first and second mood vectors to a common vector space based on the retrieved first score that quantifies the congruency between the first and second moods.
8. The method of claim 7 , wherein:
the normalizing of the first and second mood vectors to the common vector space includes modifying the first value from the first mood vector based on the retrieved first score and modifying the second value from the second mood vector based on the retrieved first score.
9. The method of claim 8 , wherein:
the generating the second score is based on the modified first value from the first mood vector and the modified second value from the second mood vector.
10. The method of claim 1 , wherein:
the first mood vector is a first portion of a first mood profile that specifies more mood-value pairs than the first mood vector, the first mood vector and the first mood profile each describing the first media data; and
the second mood vector is a second portion of a second mood profile that specifies more mood-value pairs than the second mood vector, the second mood vector and the second mood profile each describing the second media data.
11. The method of claim 10 , wherein:
the first value assigned to the first mood by the first mood-value pair is predominant among other values assigned by other mood-value pairs within the first mood vector by having a larger absolute value than the other values within the first mood vector; and
the second value assigned to the second mood by the second mood-value pair is predominant among other values assigned by other mood-value pairs within the second mood vector by having a larger absolute value than the other values within the second mood vector.
12. The method of claim 1 , wherein:
the first mood vector is a reduced representation of a first mood profile that describes the first media data and specifies more mood-value pairs than the first mood vector; and
the second mood vector is a reduced representation of a second mood profile that describes the second media data and specifies more mood-value pairs than the second mood vector.
13. The method of claim 1 , wherein:
the first mood-value pair of the first mood vector is a first subset of a first plurality of mood-value pairs included in a first mood profile that describes the first media data; and
the second mood-value pair of the second mood vector is a second subset of a second plurality of mood-value pairs included in a second mood profile that describes the second media data.
14. The method of claim 1 further comprising:
comparing the generated second score to a threshold value; and
providing an indicator that the second media data is similar to the first media data based on the comparing of the generated second score to the threshold value.
15. 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, from a database, a first mood vector that describes first media data and specifies a first mood-value pair, the first mood-value pair assigning a first value to a first mood;
accessing, from the database, a second mood vector that describes a second media data and specifies a second mood-value pair, the second mood-value pair assigning a second value to a second mood;
retrieving a first score that quantifies congruency between the first and second moods, the first score being retrieved from a data structure that correlates the first and second moods;
generating a second score that quantifies congruency between the first and second mood vectors, the generating the second score being based on the accessed first and second values and the retrieved first score; and
based on the generated second score, adding the first media data to a playlist.
16. The non-transitory machine-readable storage medium of claim 15 , wherein:
the data structure that correlates the first and second moods is a mood correlates table in which a cell that indicates a relationship between the first mood and the second mood contains the first score that quantifies the congruency between the first and second moods; and
the retrieving of the first score from the data structure retrieves the first score from the mood correlates table.
17. The non-transitory machine-readable storage medium of claim 15 , wherein:
the generating of the second score includes normalizing the first and second mood vectors to a common vector space based on the retrieved first score that quantifies the congruency between the first and second moods;
the normalizing of the first and second mood vectors to the common vector space includes modifying the first value from the first mood vector based on the retrieved first score and modifying the second value from the second mood vector based on the retrieved first score; and
the generating the second score is based on the modified first value from the first mood vector and the modified second value from the second mood vector.
18. A system comprising:
a memory that stores instructions;
a database that stores mood vectors;
one or more processors configured by the instructions to perform operations comprising:
accessing, from the database, a first mood vector that describes first media data and specifies a first mood-value pair, the first mood-value pair assigning a first value to a first mood; and
accessing, from the database, a second mood vector that describes a second media data and specifies a second mood-value pair, the second mood-value pair assigning a second value to a second mood;
retrieving a first score that quantifies congruency between the first and second moods, the first score being retrieved from a data structure that correlates the first and second moods;
generating a second score that quantifies congruency between the first and second mood vectors, the generating the second score being based on the accessed first and second values and the retrieved first score; and
based on the generated second score, adding the first media data to a playlist.
19. The system of claim 18 , wherein:
the data structure that correlates the first and second moods is a mood correlates table in which a cell that indicates a relationship between the first mood and the second mood contains the first score that quantifies the congruency between the first and second moods; and
the operations further comprise retrieving the first score from the mood correlates table.
20. The system of claim 18 , wherein the operations further comprise:
normalizing the first and second mood vectors to a common vector space by modifying the first value from the first mood vector based on the retrieved first score and modifying the second value from the second mood vector based on the retrieved first score; and
generating the second score based on the modified first value from the first mood vector and the modified second value from the second mood vector.
21. The system of claim 18 , wherein the operations further comprise:
comparing the generated second score to a threshold value; and
providing an indicator that the second media data is similar to the first media data based on a comparison of the generated second score to the threshold value.