Query response using media consumption history

1. A computer-implemented method comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data by:

identifying the particular media item based on detecting a match between one or more features of the environmental audio data associated with the content playing in the environment of the user and one or more features of the particular media item; and

determining that the particular media item is identified in a media consumption database that identifies media items that are identified as having been previously consumed by the user; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the identified previous consumption of the particular media item.

2. The computer-implemented method of claim 1 , wherein providing the response to the natural language query submitted by the user comprises:

accessing, at the media consumption database, information that specifies contextual information associated with the previous consumption of the particular media item by the user; and

providing the response to the natural language query submitted by the user that includes at least a portion of the contextual information associated with the previous consumption of the particular media item by the user.

3. A computer-implemented method comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item, wherein the natural language query submitted by the user specifies an entity type; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the previous consumption of the particular media item of the specified entity type that is determined based at least on the environmental audio data associated with the content playing in the environment of the user.

4. The computer-implemented method of claim 3 , wherein providing the response to the natural language query submitted by the user that identifies the context of the previous consumption of the particular media item of the specified entity type comprises:

obtaining a transcription of the natural language query submitted by the user;

comparing the transcription of the natural language query submitted by the user to one or more keyword phrases that are each associated with an entity type; and

selecting the entity type based on determining that the transcription of the natural language query submitted by the user contains a particular keyword phrase that is associated with the selected entity type.

5. The computer-implemented method of claim 1 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item for output in a first region of a user interface displayed at the user device of the user; and

providing information that indicates one or more characteristics of the particular media item in a second region of the user interface displayed at the user device of the user.

6. The computer-implemented method of claim 1 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item in a first region of the user device of the user; and

providing one or more search results identified in response to the natural language query submitted by the user in a second region of the user interface displayed at the user device of the user.

7. A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data by:

identifying the particular media item based on detecting a match between one or more features of the environmental audio data associated with the content playing in the environment of the user and one or more features of the particular media item; and

determining that the particular media item is identified in a media consumption database that identifies media items that are identified as having been previously consumed by the user; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the identified previous consumption of the particular media item.

8. The system of claim 7 , wherein providing the response to the natural language query submitted by the user comprises:

accessing, at the media consumption database, information that specifies contextual information associated with the previous consumption of the particular media item by the user; and

providing the response to the natural language query submitted by the user that includes at least a portion of the contextual information associated with the previous consumption of the particular media item by the user.

9. A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item, wherein the natural language query submitted by the user specifies an entity type; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the previous consumption of the particular media item of the specified entity type that is determined based at least on the environmental audio data associated with the content playing in the environment of the user.

10. The system of claim 9 , wherein providing the response to the natural language query submitted by the user that identifies the context of the previous consumption of the particular media item of the specified entity type comprises:

obtaining a transcription of the natural language query submitted by the user;

comparing the transcription of the natural language query submitted by the user to one or more keyword phrases that are each associated with an entity type; and

selecting the entity type based on determining that the transcription of the natural language query submitted by the user contains a particular keyword phrase that is associated with the selected entity type.

11. The system of claim 7 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item for output in a first region of a user interface displayed at the user device of the user; and

providing information that indicates one or more characteristics of the particular media item in a second region of the user interface displayed at the user device of the user.

12. The system of claim 7 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item in a first region of the user device of the user; and

providing one or more search results identified in response to the natural language query submitted by the user in a second region of the user interface displayed at the user device of the user.

13. A computer-readable storage device encoded with a computer program, the program comprising instructions that, if executed by one or more computers, cause the one or more computers to perform operations comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data by:

identifying the particular media item based on detecting a match between one or more features of the environmental audio data associated with the content playing in the environment of the user and one or more features of the particular media item; and

determining that the particular media item is identified in a media consumption database that identifies media items that are identified as having been previously consumed by the user; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the identified previous consumption of the particular media item.

14. The computer-readable storage device of claim 13 , wherein providing the response to the natural language query submitted by the user comprises:

accessing, at the media consumption database, information that specifies contextual information associated with the previous consumption of the particular media item by the user; and

providing the response to the natural language query submitted by the user that includes at least a portion of the contextual information associated with the previous consumption of the particular media item by the user.

15. A computer-readable storage device encoded with a computer program, the program comprising instructions that, if executed by one or more computers, cause the one or more computers to perform operations comprising:

receiving waveform data generated by a user device of a user, the waveform data comprising:

an utterance detected by the user device, the utterance corresponding to a natural language query submitted by the user that requests information relating to a context of a prior consumption of a media item, wherein the natural language query submitted by the user specifies an entity type; and

environmental audio data detected by the user device within a threshold amount of time before or after detecting the utterance, the environmental audio data associated with content playing in an environment of the user;

processing the waveform data to separate the utterance corresponding to the natural language query from the environmental audio data by:

detecting voice activity of the user in a portion of the waveform data; and

extracting the portion of the waveform data that includes the detected voice activity of the user to separate the utterance corresponding to the natural language query from the environmental audio data;

identifying a previous consumption of a particular media item based on the environmental audio data; and

providing, to the user device, a response to the natural language query submitted by the user that identifies a context of the previous consumption of the particular media item of the specified entity type that is determined based at least on the environmental audio data associated with the content playing in the environment of the user.

16. The computer-readable storage device of claim 13 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item for output in a first region of a user interface displayed at the user device of the user; and

providing information that indicates one or more characteristics of the particular media item in a second region of the user interface displayed at the user device of the user.

17. The computer-readable storage device of claim 13 , wherein providing the response to the natural language query submitted by the user comprises:

providing information that identifies the context of the previous consumption of the particular media item in a first region of the user device of the user; and

providing one or more search results identified in response to the natural language query submitted by the user in a second region of the user interface displayed at the user device of the user.