Advertisement selection based on demographic information inferred from media item preferences

1. A computer-implemented method for advertisement selection based on demographic information inferred from media item preferences, comprising:

establishing a set of baseline users with baseline user profiles providing baseline demographic information about the baseline users and correlating the baseline users with preferences for musicological features of media items;

monitoring media preferences of a user associated with a partial user profile specifying a subset of the baseline demographic information about the user;

deconstructing the monitored media preferences to identify preferences of the user for musicological features of media items, the deconstructing comprising:

identifying a plurality of media items associated with the monitored media preferences,

determining a feedback score for each of the media items,

identifying musicological features that characterize the media items using a database of features and a mapping of media items to specific musicological features, the musicological features describing underlying characteristics of the media items,

applying the feedback scores for the media items to their respective identified musicological features, and

aggregating the individual musicological features and the musicological features' associated feedback scores such that the media preferences of the user are represented as a set of individual musicological features with associated scores representing the user's likes and dislikes;

determining whether the preferences of the user for musicological features of the media items are associated with a threshold number of the baseline users;

responsive to the preferences of the user for musicological features of the media items not being associated with the threshold number of the plurality of baseline users:

querying the user to provide the baseline demographic information, and

adding the user to the set of baseline users responsive to receiving the queried-for baseline demographic information; and

responsive to the preferences of the user for musicological features of the media items being associated with the threshold number of the plurality of baseline users:

inferring profile information missing from the partial user profile responsive to the preferences of the user for musicological features of the media items and the baseline user profiles;

selecting an advertisement using the inferred profile information, and

providing the selected advertisement to a client device used by the user.

2. The method of claim 1 , wherein inferring profile information missing from the partial user profile responsive to the preferences of the user for musicological features of the media items, comprises:

using the learned relationships between demographic information and media feature preferences in combination with the preferences of the user for musicological features of the media items, to infer expanded demographic information about the user that is missing from the partial user profile of the user; and

augmenting the partial user profile with the some of the retrieved expanded demographic information.

3. The method of claim 1 , wherein partial demographic information about the user is selected from the group consisting of an age of the user, location of the user, and gender of the user.

4. The method of claim 1 , further comprising:

establishing a set of users that are each associated with respective expanded demographic information;

monitoring media preferences of the set of users;

deconstructing the monitored media preferences of the set of users into media features;

learning correlations between the expanded demographic information and the deconstructed monitored media preferences; and

storing the learned mappings as the baseline information.

5. The method of claim 4 , wherein learning correlations between the expanded demographic information and the deconstructed monitored media preferences comprises:

using machine learning to identify relationships between expanded demographic information of the baseline users and media feature preferences of those users; and

storing the learned mappings as baseline information.

6. The method of claim 1 , wherein selecting the advertisement using the inferred profile information comprises:

retrieving demographic information from the partial user profile that includes the inferred profile information;

extracting expanded demographic information from the retrieved partial user profile;

comparing a portion of the expanded demographic information to targeting information associated with different candidate advertisements; and

selecting a candidate advertisement that has a threshold percentage of targeting information that matches the portion of the expanded demographic information.

7. A non-transitory computer-readable storage medium storing executable computer program instructions for advertisement selection based on demographic information inferred from media item preferences, the instructions executable to perform steps comprising:

establishing a set of baseline users with baseline user profiles providing baseline demographic information about the baseline users and correlating the baseline users with preferences for musicological features of media items;

monitoring media preferences of a user associated with a partial user profile specifying a subset of the baseline demographic information about the user;

deconstructing the monitored media preferences to identify preferences of the user for musicological features of media items, the deconstructing comprising:

identifying a plurality of media items associated with the monitored media preferences,

determining a feedback score for each of the media items,

identifying musicological features that characterize the media items using a database of features and a mapping of media items to specific musicological features, the musicological features describing underlying characteristics of the media items,

applying the feedback scores for the media items to their respective identified musicological features, and

aggregating the individual musicological features and the musicological features' associated feedback scores such that the media preferences of the user are represented as a set of individual musicological features with associated scores representing the user's likes and dislikes;

determining whether the preferences of the user for musicological features of the media items are associated with a threshold number of the baseline users;

responsive to the preferences of the user for musicological features of the media items not being associated with the threshold number of the plurality of baseline users:

querying the user to provide the baseline demographic information, and

adding the user to the set of baseline users responsive to receiving the queried-for baseline demographic information; and

responsive to the preferences of the user for musicological features of the media items being associated with the threshold number of the plurality of baseline users:

inferring profile information missing from the partial user profile responsive to the preferences of the user for musicological features of the media items and the baseline user profile;

selecting an advertisement using the inferred profile information, and

providing the selected advertisement to a client device used by the user.

8. The computer-readable medium of claim 7 , wherein inferring profile information missing from the partial user profile responsive to the preferences of the user for musicological features of the media items, comprises:

using the learned relationships between demographic information and media feature preferences in combination with the preferences of the user for musicological features of the media items, to infer expanded demographic information about the user that is missing from the partial user profile of the user; and

augmenting the partial user profile with the some of the retrieved expanded demographic information.

9. The computer-readable medium of claim 7 , wherein partial demographic information about the user is selected from the group consisting of an age of the user, location of the user, and gender of the user.

10. The computer-readable medium of claim 7 , further comprising:

establishing a set of users that are each associated with respective expanded demographic information;

monitoring media preferences of the set of users;

deconstructing the monitored media preferences of the set of users into media features;

learning correlations between the expanded demographic information and the deconstructed monitored media preferences; and

storing the learned mappings as the baseline information.

11. The computer-readable medium of claim 10 , wherein learning correlations between the expanded demographic information and the deconstructed monitored media preferences comprises:

using machine learning to identify relationships between expanded demographic information of the baseline users and media feature preferences of those users; and

storing the learned mappings as baseline information.

12. The computer-readable medium of claim 7 , wherein selecting the advertisement using the inferred profile information comprises:

retrieving demographic information from the partial user profile that includes the inferred profile information;

extracting expanded demographic information from the retrieved partial user profile;

comparing a portion of the expanded demographic information to targeting information associated with different candidate advertisements; and

selecting a candidate advertisement that has a threshold percentage of targeting information that matches the portion of the expanded demographic information.

13. A device for advertisement selection based on demographic information inferred from media item preferences, comprising:

a processor configured to execute modules; and

a memory storing the modules, the modules comprising:

a baseline module configured to establish a set of baseline users with baseline user profiles providing baseline demographic information about the baseline users and correlating the baseline users with preferences for musicological features of media items;

a monitoring module configured to monitor media preferences of a user associated with a partial user profile specifying a subset of the baseline demographic information about the user,

a deconstruction module configured to deconstruct the monitored media preferences to identify preferences of the user for musicological features of the media items, the deconstruction module further configured to:

identify a plurality of media items associated with the monitored media preferences,

determine a feedback score for each of the media items,

identify musicological features that characterize the media items using a database of features and a mapping of media items to specific musicological features, the musicological features describing underlying characteristics of the media items,

apply the feedback scores for the media items to their respective identified musicological features, and

aggregate the individual musicological features and the musicological features' associated feedback scores such that the media preferences of the user are represented as a set of individual musicological features with associated scores representing the user's likes and dislikes;

wherein the baseline module is further configured to:

determine whether the preferences of the user for musicological features of the media items are associated with a threshold number of the baseline users, and

responsive to the preferences of the user for musicological features of the media items not being associated with the threshold number of the plurality of baseline users:

query the user to provide the baseline demographic information, and

add the user to the set of baseline users responsive to receiving the queried-for baseline demographic information; and

responsive to the preferences of the user for musicological features of the media items being associated with the threshold number of the plurality of baseline users:

an inference module configured to infer profile information missing from the partial user profile responsive to the preferences of the user for musicological features of the media items and the baseline user profiles,

a content selection module configured to select an advertisement using the inferred profile information, and

an advertisement server module configured to provide the selected advertisement to a client device used by the user.

14. The device of claim 13 , wherein the inference module is further configured to:

use the learned relationships between demographic information and media feature preferences in combination with the preferences of the user for musicological features of the media items, to infer expanded demographic information about the user that is missing from the partial user profile of the user; and

augment the partial user profile with the some of the retrieved expanded demographic information.

15. The device of claim 13 , wherein the baseline module is configured to establish a set of users that are each associated with respective expanded demographic information; and

wherein,

the monitoring module is further configured to monitor media preferences of the set of users,

the deconstruction module is further configured to deconstruct the monitored media preferences of the set of users into media features,

the correlation module is further configured to:

learn correlations between the expanded demographic information and the deconstructed monitored media preferences; and

store the learned mappings as the baseline information.

16. The device of claim 15 , wherein the correlation module is further configured to:

use machine learning to identify relationships between expanded demographic information of the baseline users and media feature preferences of those users; and

store the learned mappings as baseline information.

17. The device of claim 16 , wherein the advertisement selection module is further configured to:

retrieve demographic information from the partial user profile that includes the inferred profile information;

extract expanded demographic information from the retrieved partial user profile;

compare a portion of the expanded demographic information to targeting information associated with different candidate advertisements; and

select a candidate advertisement that has a threshold percentage of targeting information that matches the portion of the expanded demographic information.