Cross media recommendation

1. A system for cross-media recommendation, comprising:

a first database operable to store a plurality of user taste profiles corresponding to a first domain;

a second database operable to store a plurality of media item vectors corresponding to a second domain; and

at least one processor configured to:

generate a plurality of weighted term vectors by multiplying, for each term in a user taste profile of the plurality of user taste profiles, an affinity by a probability that the term is associated with a media item;

generate vector quantized media data by vector quantizing the plurality of media item vectors;

map the weighted term vectors to the vector quantized media data to form relationships between the user taste profiles and one or more media items;

train a model based on the map to learn the relationships between the user taste profiles and the one or more media items; and

apply an evaluation taste profile in the first domain to the model to obtain a media item corresponding to the second domain.

2. The system according to claim 1 , wherein the first domain is music and the second domain is any one, or a combination, of books, movies, or games.

3. The system according to claim 1 , wherein the at least one processor is further configured to:

generate a training set based on the plurality of user taste profiles and the plurality of media item vectors, wherein at least a portion of the training set includes ground truths across the first domain and the second domain; and

train the model with the training set.

4. The system according to claim 1 , wherein the at least one processor is further configured to:

obtain a confidence interval indicating an amount of confidence in a relation between the evaluation taste profile and the media item vector of the media item corresponding to the second domain.

5. A computer-implemented method for cross-media recommendation, comprising:

storing a plurality of user taste profiles corresponding to a first domain;

storing a plurality of media item vectors corresponding to a second domain;

generating a plurality of weighted term vectors by multiplying, for each term in a user taste profile of the plurality of user taste profiles, an affinity by a probability that the term is associated with a media item;

generating vector quantized media data by vector quantizing the plurality of media item vectors;

mapping the weighted term vectors to the vector quantized media data to form relationships between the user taste profiles and one or more media items;

training a model based on the map to learn the relationships between the user taste profiles and the one or more media items; and

applying an evaluation taste profile in the first domain to the model to obtain a media item corresponding to the second domain.

6. The computer-implemented method according to claim 5 , wherein the first domain is music and the second domain is any one, or a combination, of books, movies, or games.

7. The computer-implemented method according to claim 5 , further comprising the step of:

generating a training set based on the plurality of user taste profiles and the plurality of media item vectors, wherein at least a portion of the training set includes ground truths across the first domain and second domain; and

training the model with the training set.

8. The computer-implemented method according to claim 5 , further comprising the step of:

obtaining a confidence interval indicating an amount of confidence in a relation between the evaluation taste profile and the media item vector of the media item corresponding to the second domain.

9. A non-transitory computer-readable medium having stored thereon one or more sequences of instructions for causing one or more processors to perform:

storing a plurality of user taste profiles corresponding to a first domain;

storing a plurality of media item vectors corresponding to a second domain;

generating a plurality of weighted term vectors by multiplying, for each term in a user taste profile of the plurality of user taste profiles, an affinity by a probability that the term is associated with a media item;

generating vector quantized media data by vector quantizing the plurality of media item vectors;

mapping the weighted term vectors to the vector quantized media data to form relationships between the user taste profiles and one or more media items;

training a model based on the map to learn the relationships between the user taste profiles and the one or more media items; and

applying an evaluation taste profile in the first domain to the model to obtain a media item corresponding to the second domain.

10. The non-transitory computer-readable medium of claim 9 , wherein the first domain is music and the second domain is any one, or a combination, of books, movies, or games.

11. The non-transitory computer-readable medium of claim 9 , further having stored thereon a sequence of instructions for causing the one or more processors to perform:

generating a training set based on the plurality of user taste profiles and the plurality of media item vectors, wherein at least a portion of the training set includes ground truths across the first domain and second domain; and

training the model with the training set.

12. The non-transitory computer-readable medium of claim 9 , further having stored thereon a sequence of instructions for causing the one or more processors to perform:

obtaining a confidence interval indicating an amount of confidence in a relation between the evaluation taste profile and the media item vector of the media item corresponding to the second domain.