Cross media recommendation

Patent metadata

Field	Value
Publication number	US-9613118-B2
Application number	US-201414213749-A
Country	US
Kind code	B2
Filing date	Mar 14, 2014
Priority date	Mar 18, 2013
Publication date	Apr 4, 2017
Grant date	Apr 4, 2017

Abstract

Methods, systems and computer program products are provided for cross-media recommendation by store a plurality of taste profiles corresponding to a first domain and a plurality of media item vectors corresponding to a second domain. An evaluation taste profile in the first domain is applied to a plurality of models that have been generated based on relationship among the plurality of taste profiles and the plurality of media item vectors, and obtain a plurality of resulting codes corresponding to at least one of the plurality of media item vectors in the second domain.

First claim

What is claimed is:

1. A system for cross-media recommendation, comprising:
a first database operable to store a plurality of 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 training set based on the plurality of taste profiles and the plurality of media item vectors;
apply an evaluation taste profile in the first domain to a plurality of models generated based on a relationship among the plurality of taste profiles and the plurality of media item vectors, wherein the plurality of models are trained based on the training set;
obtain a plurality of resulting codes corresponding to at least one of the plurality of media item vectors in the second domain;
generate a plurality of weighted term vectors based on the plurality of taste profiles;
generate vector quantized media data by vector quantizing the plurality of media item vectors; and
generate a map of the weighted term vectors to the vector quantized media data,
wherein the plurality of weighted term vectors are generated by multiplying, for each term in a taste profile, an affinity by a probability that the term is associated with a media item.

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 plurality of media item vectors are vector quantized by applying the plurality of media item vectors to a k-means clustering algorithm.

4. 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 taste profiles corresponding to a first domain;
storing a plurality of media item vectors corresponding to a second domain;
generating a training set based on the plurality of taste profiles and the plurality of media item vectors;
applying an evaluation taste profile in the first domain to a plurality of models generated based on a relationship among the plurality of taste profiles and the plurality of media item vectors, wherein the plurality of models are trained based on the training set;
obtaining a plurality of resulting codes corresponding to at least one of the plurality of media item vectors in the second domain;
generating a plurality of weighted term vectors based on the plurality of taste profiles;
generating vector quantized media data by vector quantizing the plurality of media item vectors; and
generating a map of the weighted term vectors to the vector quantized media data,
wherein the weighted term vector is generated by multiplying, for each term in a taste profile, an affinity by a probability that the term is associated with a media item.

5. The computer-readable medium of claim 4 , wherein the first domain is music and the second domain is any one, or a combination, of books, movies, or games.

6. The computer-readable medium of claim 4 , wherein the plurality of media item vectors are vector quantized by applying the plurality of media item vectors to a k-means clustering algorithm.

7. A method for cross-media recommendation, comprising the steps of:
storing a plurality of taste profiles corresponding to a first domain;
storing a plurality of media item vectors corresponding to a second domain;
generating a training set based on the plurality of taste profiles and the plurality of media item vectors;
applying an evaluation taste profile in the first domain to a plurality of models generated based on a relationship among the plurality of taste profiles and the plurality of media item vectors, wherein the plurality of models are trained based on the training set;
obtaining a plurality of resulting codes corresponding to at least one of the plurality of media item vectors in the second domain;
generating a plurality of weighted term vectors based on the plurality of taste profiles;
generating vector quantized media data by vector quantizing the plurality of media item vectors; and
generating a map of the weighted term vectors to the vector quantized media data,
wherein the weighted term vector is generated by multiplying, for each term in a taste profile, an affinity by a probability that the term is associated with a media item.

8. The method according to claim 7 , wherein the first domain is music and the second domain is any one, or a combination, of books, movies, or games.

9. The method according to claim 7 , wherein the plurality of media item vectors are vector quantized by applying the plurality of media item vectors to a k-means clustering algorithm.

Assignees

• The Echo Nest Corp
• Spotify Ab

Inventors

• Whitman Brian

Classifications

• G06F16/435

  Filtering based on additional data, e.g. user or group profiles · CPC title

• G06F16/735

  Filtering based on additional data, e.g. user or group profiles · CPC title

• G06F16/24578

  using ranking · CPC title

• G06N5/025

  Extracting rules from data · CPC title

• G06F16/26 Primary

  Visual data mining; Browsing structured data · CPC title