Content analysis to enhance voice search

1. A method comprising:

based on a determination that a first media content item describes a second media content item, determining, by a computing device via an analysis of the first media content item, a plurality of keywords corresponding to the first media content item;

associating one or more keywords of the plurality of keywords with the second media content item;

receiving a search request indicative of a keyword of the one or more keywords;

performing a search for media content items associated with the keyword of the one or more keywords; and

outputting, for display, an indication corresponding to the second media content item as a result of the search.

2. The method of claim 1 , wherein the first media content item is an advertisement comprising content describing the second media content item.

3. The method of claim 1 , wherein the determining the plurality of keywords comprises at least one of:

determining the plurality of keywords from an audio signal of the first media content item by performing speech recognition on the audio signal,

determining the plurality of keywords by parsing closed caption information associated with the first media content item,

determining the plurality of keywords by parsing metadata of the first media content item, or

determining the plurality of keywords from a video signal of the first media content item by performing optical character recognition on video frames of the video signal.

4. The method of claim 1 , wherein the associating the one or more keywords with the second media content item comprises:

clustering at least two keywords of the plurality of keywords to form an n-gram; and

associating the n-gram with metadata of the second media content item, wherein the metadata is configured for use during the search to locate the second media content item.

5. The method of claim 1 , wherein the associating the one or more keywords with the second media content item comprises:

adding the one or more keywords to metadata associated with the second media content item, wherein the metadata is configured for use during the search to locate the second media content item.

6. The method of claim 1 , further comprising:

monitoring user input for a word describing the second media content item; and

adding the word to metadata associated with the second media content item.

7. The method of claim 1 , wherein the receiving the search request comprises:

receiving a speech input comprising one or more user utterances; and

generating a cluster of user utterances from the one or more user utterances, wherein the performing the search comprises comparing the cluster of user utterances with the one or more keywords.

8. The method of claim 7 , wherein the associating the one or more keywords with the second media content item comprises:

clustering the one or more keywords into one or more trigrams; and

associating the one or more trigrams with the second media content item, wherein the comparing the cluster of user utterances with the one or more keywords comprises comparing the cluster of user utterances with the one or more trigrams.

9. The method of claim 1 , further comprising:

clustering the one or more keywords into one or more keyword trigrams, wherein the receiving the search request comprises:

receiving a speech input comprising one or more user utterances; and

generating one or more user utterance trigrams from the one or more user utterances, and wherein the performing the search comprises:

comparing the one or more user utterance trigrams with the one or more keyword trigrams.

10. The method of claim 7 , the method further comprising:

determining that the cluster of user utterances comprises an incorrectly spelled utterance that is homophonous to a keyword describing the second media content item; and

before performing the search, correcting a spelling of the incorrectly spelled utterance.

11. A method comprising:

determining, by a computing device, an advertisement comprising media content describing a scheduled media item;

determining, based on the media content, one or more keywords describing the scheduled media item;

associating the one or more keywords with the scheduled media item;

receiving a search input comprising one or more user utterances;

determining that the scheduled media item is a search result by comparing the one or more user utterances with the one or more keywords; and

outputting, for display, an indication corresponding to the scheduled media item.

12. The method of claim 11 , wherein the determining the one or more keywords describing the scheduled media item comprises:

determining, based on the media content, at least one online database entry related to the scheduled media item; and

determining the one or more keywords from the at least one online database entry.

13. The method of claim 11 , wherein the associating the one or more keywords with the scheduled media item comprises:

adding the one or more keywords to metadata associated with the scheduled media item, wherein the metadata is configured for use during a search to locate the scheduled media item.

14. The method of claim 11 , further comprising:

generating a cluster of user utterances from the one or more user utterances, wherein the determining that the scheduled media item is the search result is based on a determination that the one or more keywords corresponds to the cluster of user utterances.

15. The method of claim 11 , wherein the associating the one or more keywords with the scheduled media item comprises:

clustering the one or more keywords into one or more trigrams; and

associating each trigram of the one or more trigrams with the scheduled media item, wherein the determining that the scheduled media item is the search result comprises determining that at least one trigram of the one or more trigrams matches at least one user utterance of the one or more user utterances.

16. The method of claim 15 , wherein the determining that the at least one trigram matches the at least one user utterance comprises:

generating one or more user utterance trigrams from terms in the at least one user utterance; and

comparing the one or more user utterance trigrams with the at least one trigram of the one or more trigrams.

17. A method comprising:

determining, by a computing device via an analysis of a first media content item, a plurality of keywords describing a second media content item;

clustering the plurality of keywords into a plurality of trigrams;

associating the plurality of trigrams with the second media content item;

receiving a search input comprising one or more user utterances;

determining that the second media content item is a search result, based on a determination that the one or more user utterances corresponds to at least one trigram of the plurality of trigrams; and

outputting, for display, an indication corresponding to the second media content item.

18. The method of claim 17 , further comprising:

generating one or more user utterance trigrams from terms in the one or more user utterances, wherein the determining that the second media content item is the search result comprises comparing the one or more user utterance trigrams with the at least one trigram of the plurality of trigrams.

19. The method of claim 17 , wherein the determining the plurality of keywords describing the second media content item comprises:

determining, based on the analysis of the first media content item, at least one online database entry related to the second media content item; and

determining at least one keyword of the plurality of keywords based on the at least one online database entry.

20. The method of claim 17 , wherein the first media content item is an advertisement comprising media content that describes the second media content item.