Audiovisual content item transcript search engine
Self-learning systems process data in real-time and output the processed data to client applications in an effective manner. They comprise a capture platform that captures data and generates a stream of text, a text decoding server that extracts individual words from the stream of text, an entity extractor that identifies entities, a trending engine that outputs trending results, and a live queue broker that filters the trending results. The self-learning systems provide more efficient realization of Boxfish technologies, and provide or work in conjunction with real-time processing, storage, indexing, and delivery of segmented video. Furthermore, the self-learning systems efficiently perform entity relationing by creating entity network graphs, and are operable to identify advertisements from the data.
1. A method comprising:
receiving, by a computing device, a first search term;
identifying by the computing device, one or more second search terms relating to the first search term;
determining a plurality of audiovisual content items in which the first search term is spoken;
causing output, at a first time, of a plurality of image tiles corresponding to the plurality of audiovisual content items, wherein a first image tile of the plurality of image tiles comprises:
a thumbnail image associated with a first audiovisual content item of the plurality of audiovisual content items; and
a first transcript portion associated with the first audiovisual content item and comprising the first search term;
detecting, after the causing the output of the plurality of image tiles, an occurrence of the one or more second search terms being spoken in the first audiovisual content item at a second time later than the first time by monitoring, during output of the plurality of image tiles, a transcript of the first audiovisual content item;
based on the receiving the first search term, and based on detecting the occurrence of the one or more second search terms after the causing the output of the plurality of image tiles, replacing, during output of the plurality of image tiles, the first transcript portion with a second transcript portion that comprises the one or more second search terms;
receiving, from a user device, a selection of the first image tile; and
causing, via a display, playback of the first audiovisual content item, wherein the playback begins at a playback time corresponding to the new-occurrence of the one or more second search terms in the first audiovisual content item.
2. The method of claim 1 , wherein the replacing the first transcript portion is further based on one or more of:
entities corresponding to the first search term,
co-occurrences of the entities in a database, or
a different mention of the first search term in a different content item.
3. The method of claim 1 , wherein the replacing the first transcript portion comprises determining trending results in a particular geographic region.
4. The method of claim 1 , wherein the replacing the first transcript portion comprises determining trending results corresponding to a particular genre.
5. The method of claim 1 , wherein the replacing the first transcript portion with the second transcript portion comprises:
determining entities corresponding to the first search term by:
extracting individual words from a database;
determining that a plurality of different words of the individual words relate to the same entity; and
normalizing a plurality of entities respectively associated with the plurality of different words to comprise a common representation of the same entity; and
determining a frequency of mentions of each of the normalized plurality of entities.
6. The method of claim 1 , wherein the replacing the first transcript portion comprises determining trending results corresponding to a source of content item.
7. The method of claim 1 , wherein the replacing the first transcript portion comprises:
determining entities corresponding to the first search term; and
determining trending results based on how frequently each entity of the entities occurs in a database by comparing a percentage of change of a plurality of phrases mentioned in the database against historic frequency metrics for each of the plurality of phrases.
8. The method of claim 1 , further comprising:
causing, based on a frequency of mentions of the first search term, the user device to output an advertisement.
9. The method of claim 1 , further comprising:
recognizing an advertisement in a database; and
filtering the advertisement from being outputted in the plurality of audiovisual content items in which the first search term is spoken.
10. The method of claim 1 , comprising:
storing, in a database, frequency counts of how often any two entities co-occur in one or more of the plurality of audiovisual content items, wherein the frequency counts are based on a time and a date of each co-occurrence, wherein the replacing the first transcript portion comprises querying the database for the first search term.
11. The method of claim 1 , wherein the replacing the first transcript portion comprises comparing the first search term to a threshold.
12. The method of claim 1 , wherein each image tile, of the plurality of image tiles, further comprises a time elapsed since the search term was mentioned in a corresponding audiovisual content item.
13. The method of claim 1 , wherein the first audiovisual content item is a live content item, and wherein the occurrence of the one or more second search terms is detected after the causing output of the plurality of image tiles.
14. The method of claim 1 , further comprising:
receiving, at a third time, a user request to be notified of future occurrences of the first search term being spoken in any of the plurality of audiovisual content items;
determining, during the playback of the first audiovisual content item, an occurrence of the first search term being spoken in a second audiovisual content item of the plurality of audiovisual content items at a fourth time later than the third time; and
outputting a notification that:
partially overlays a portion of the first audiovisual content item;
indicates the search term; and
comprises a second thumbnail image corresponding to the second audiovisual content item.
15. The method of claim 1 , wherein the first image tile further comprises an indication of a time of a new occurrence of the first search term.
16. The method of claim 1 , replacing the thumbnail image with a second thumbnail image associated with the new-occurrence of the one or more second search terms in the first audiovisual content item.
17. The method of claim 1 , wherein the first audiovisual content item is a live audiovisual content item, and wherein the method further comprises continuing to update the first image tile as the first search term is detected in a transcript corresponding to the live audiovisual content item.
18. The method of claim 1 , further comprising updating the first image tile periodically at predetermined time intervals.
19. A method comprising,
receiving, by a computing device, a user preference related to a topic by receiving a first search term;
determining, based on the topic, a plurality of words comprising the first search term by identifying one or more second search terms relating to the first search term;
causing, at a first time and based on detection of at least one first word of the plurality of words spoken in a plurality of content items, output of one or more image tiles corresponding to one or more mentions of the topic in the plurality of content wherein a first image tile of the one or more image tiles comprises:
a thumbnail image associated with a first content item of the plurality of content items; and
a first transcript portion associated with the first content item and comprising the at least one first word; and
based on the receiving the user preference related to the topic, and after causing output of the one or more image tiles, continuing to monitor for additional detection of any one of the plurality of words spoken in the plurality of content items by monitoring, during output of the one or more image tiles, one or more transcripts of the plurality of content items; and
replacing, based on a most recent detection of at least one second word of the plurality of words spoken in the first content item at a second time after the first time, and based on the receiving the user preference related to the topic, the first transcript portion with a second transcript portion that comprises the at least one second word of the plurality of words.
20. The method of claim 19 , further comprising:
outputting, based on a mention of the topic in a second content item of the plurality of content items, a notification that comprises an option to view the second content item, wherein a selection of the option causes output of the second content item.
21. The method of claim 19 , wherein the plurality of content items comprises a plurality of video programs.
22. The method of claim 19 , wherein the first image tile further comprises a time elapsed since the most recent detection of the at least one second word in the first content item.
23. The method of claim 19 , further comprising:
periodically updating at least one of the one or more image tiles.
24. The method of claim 19 , further comprising:
during playback of the first content item, based on determining an occurrence of one or more of the plurality of words in a second content item of the plurality of content items, outputting a notification that partially overlays a portion of the first content item and indicates that the topic was mentioned in the second content item.
25. The method of claim 19 , wherein the first image tile further comprises:
a time of a most recent mention of the topic.
26. The method of claim 19 , further comprising replacing the thumbnail image with a second thumbnail image associated with the most recently detected one or more of the plurality of words.
27. A method comprising:
receiving, by a computing device, a first search term;
identifying, by the computing device, one or more second search terms relating to the first search term;
outputting a listing of a plurality of image tiles corresponding to a plurality of audiovisual content items in which the first search term is spoken, wherein a first image tile of the plurality of image tiles comprises:
a thumbnail image associated with the-a first audiovisual content item of the plurality of audiovisual content items; and
a transcript portion associated with the first audiovisual content item and comprising the first search term;
receiving, at a first time, a user request to be notified of future occurrences of the first search term being spoken in any of the plurality of audiovisual content items;
monitoring, based on the user request and during output of the listing of the plurality of image tiles, changes to transcripts of the plurality of audiovisual content items; and
based on receiving a selection of the first image tile in the listing:
causing output of the first audiovisual content item;
determining, based on the monitoring, during the output of the listing of the plurality of image tiles, based on the changes to the transcripts, and based on the user request, an occurrence of the one or more second search terms being spoken in a second audiovisual content item, of the plurality of audiovisual content items, at a second time later than the first time; and
based on the user request, based on the occurrence of the one or more second search terms being spoken in the second audiovisual content item, and based on receiving the user request, outputting a notification that partially overlays a portion of the first audiovisual content item and indicates that the one or more second search terms were spoken in the second audiovisual content item.
28. The method of claim 27 , wherein the notification that partially overlays the portion of the first audiovisual content item comprises an option to output the second audiovisual content item.
29. The method of claim 28 , further comprising:
causing, based on a selection of the option, replacement of output of the first audiovisual content item with output of the second audiovisual content item.
30. The method of claim 27 , wherein the notification that partially overlays the portion of the first audiovisual content item comprises a user interface element for recording the second audiovisual content item.
31. The method of claim 27 , wherein outputting the listing of the plurality of image tiles comprises:
updating, based on the occurrence, a second image tile corresponding to the second audiovisual content item.
32. The method of claim 27 , wherein the first image tile further comprises:
a time of a most recent occurrence of the first search term in the first audiovisual content item.