Topic relevance detection using automated speech recognition
A conference system automatically detects a topic in a discussion between two or more participants in a conference based on a real-time transcription of an audio component of the conference. The conference system determines that the discussion is a side conversation based on a determination that the topic is not related to any discussion points of the conference. The conference system determines which participants are related to the side conversation and schedules a future conference between these participants. The conference system generates one or more discussion points for the future conference based on the topic.
1. A method comprising:
generating a real-time transcription of a conference;
detecting a topic based on the real-time transcription;
determining that the topic is unrelated to discussion points of the conference;
determining participants related to the topic;
transmitting a prompt to the participants related to the topic to confirm that the topic is unrelated to the discussion points of the conference;
generating a future discussion point based on the topic responsive to receiving one or more confirmation responses, the future discussion point related to a future conference; and
automatically adding the future conference to respective calendars of the participants as calendar entries based on an availability of the participants, wherein the calendar entries include the future discussion point and an audio portion of the conference associated with the future discussion point.
2. The method of claim 1 , wherein detecting the topic includes processing the real-time transcription using a machine learning (ML) model trained for contextual awareness.
3. The method of claim 1 , comprising:
detecting the topic by a keyword that references one or more subjects.
4. The method of claim 1 , comprising:
detecting a phrase based on the real-time transcription and discussion points related to the conference.
5. The method of claim 1 , wherein detecting the topic includes processing the real-time transcription using a machine learning (ML) model trained for contextual awareness, the method comprising:
detecting a phrase based on the real-time transcription and discussion points related to the conference; and
detecting the topic based on a determination of keywords within a neighboring word range of the phrase.
6. The method of claim 1 , wherein detecting the topic includes performing a semantic analysis on the real-time transcription.
7. The method of claim 1 , wherein detecting the topic includes performing a semantic analysis on the real-time transcription when a duration of time that a keyword or phrase is not detected meets a threshold.
8. The method of claim 1 , wherein the calendar entries include a video portion of the conference associated with the future discussion point.
9. A server comprising:
one or more processors configured to:
generate a real-time transcription of a conference;
detect a topic based on the real-time transcription;
determine that the topic is unrelated to discussion points of the conference;
determine participants related to the topic;
transmit a prompt to the participants related to the topic to confirm that the topic is unrelated to the discussion points of the conference;
generate a future discussion point based on the topic responsive to reception of one or more confirmation responses, the future discussion point related to a future conference; and
automatically add the future conference to respective calendars of the participants as calendar entries based on an availability of the participants, wherein the calendar entries include the future discussion point and an audio portion of the conference associated with the future discussion point.
10. The server of claim 9 , wherein the one or more processors are configured to process the real-time transcription using a machine learning (ML) model trained for contextual awareness.
11. The server of claim 9 , wherein the one or more processors are configured to detect the topic by a keyword that references one or more subjects.
12. The server of claim 9 , wherein the one or more processors are configured to detect a phrase based on the real-time transcription and discussion points related to the conference.
13. The server of claim 9 , wherein the one or more processors are configured to:
detect a phrase based on the real-time transcription and discussion points related to the conference; and
detect the topic based on a determination of keywords within a neighboring word range of the phrase.
14. The server of claim 9 , wherein the one or more processors are configured to perform a semantic analysis on the real-time transcription to detect the topic.
15. A non-transitory computer-readable medium comprising instructions stored on a memory, that when executed by a processor, cause the processor to:
generate a real-time transcription of a conference;
detect a topic based on the real-time transcription;
determine that the topic is unrelated to discussion points of the conference;
determine participants related to the topic;
transmit a prompt to the participants related to the topic to confirm that the topic is unrelated to the discussion points of the conference;
generate a future discussion point based on the topic responsive to reception of one or more confirmation responses, the future discussion point related to a future conference; and
automatically add the future conference to respective calendars of the participants as calendar entries based on an availability of the participants, wherein the calendar entries include the future discussion point and an audio portion of the conference associated with the future discussion point.
16. The non-transitory computer-readable medium of claim 15 , wherein the instructions, when executed by the processor, cause the processor to process the real-time transcription using a machine learning (ML) model trained for contextual awareness.
17. The non-transitory computer-readable medium of claim 15 , wherein the instructions, when executed by the processor, cause the processor to detect the topic by a keyword that references one or more subjects.
18. The non-transitory computer-readable medium of claim 15 , wherein the instructions, when executed by the processor, cause the processor to detect a phrase based on the real-time transcription and discussion points related to the conference.
19. The non-transitory computer-readable medium of claim 15 , wherein the instructions, when executed by the processor, cause the processor to:
detect a phrase based on the real-time transcription and discussion points related to the conference; and
detect the topic based on a determination of keywords within a neighboring word range of the phrase.
20. The non-transitory computer-readable medium of claim 15 , wherein the calendar entries include a video portion of the conference.