Speech Analytics-Transcripts | C2C Community
Solved

Speech Analytics-Transcripts

  • 20 September 2023
  • 4 replies
  • 56 views

Userlevel 3
Badge +2

Does Google offer a solution that performs speech analytics on transcribed text that supports API calls. The solution will interview transcripts and provide sentiment analysis and insight on the users conversation (stutters, “Uhm”, seems unprepared). Generate a report and then graph it. 


 

icon

Best answer by malamin 20 September 2023, 22:05

View original

4 replies

Userlevel 7
Badge +35

Hello @ChambiarGirl,
Yes, Google offers a solution called the Google Cloud Speech-to-Text API that can transcribe speech into text. While the API itself does not provide built-in sentiment analysis or insights on the user's conversation, you can integrate it with other Google Cloud services to achieve the desired functionality.

To perform sentiment analysis on transcribed text, you can use the Google Cloud GenAi, VertextAi and Google Cloud Natural Language API. This API analyzes the sentiment of text and provides insights such as positive, negative, or neutral sentiment. You can extract sentiment information from the transcribed text and generate insights based on the user's conversation.

To identify stutters, "Uhm", or instances where the user seems unprepared, you can leverage various techniques such as text processing, pattern matching, or custom machine learning models. You can develop your own logic or use third-party libraries to identify and analyze these patterns within the transcribed text.

Once you have extracted sentiment analysis and insights from the transcribed text, you can generate a report summarizing the findings. You can use tools like Google Data Studio or other data visualization libraries to create graphs and visualizations based on the extracted data.

By combining the Google Cloud Speech-to-Text API, GenAi, VertextAi and Google Cloud Natural Language API, and custom logic for identifying speech patterns, you can build a comprehensive solution that performs speech analytics on transcribed text, provides sentiment analysis, generates reports, and visualizes data through graphs.

https://cloud.google.com/speech-to-text#all-features

https://cloud.google.com/speech-to-text/docs/transcribe-api

https://cloud.google.com/vertex-ai/docs/samples/aiplatform-predict-text-sentiment-analysis-sample

https://cloud.google.com/vertex-ai/docs/text-data/sentiment-analysis/train-model

Userlevel 3
Badge +2

Thank you so much. I will share with our developers. 

Userlevel 7
Badge +29

@ChambiarGirl Hi again Yvonne, 

Just like in your other topic, let me say this again, don’t forget to let us know how helpful @malamin’s answer was to you :)

Also,I noticed you haven’t introduced yourself yet in the C2C Lounge. Can you take a few minutes and do so, by using this template as a guide? It would be great for our community members to get to know you a bit better :)

 

I will be looking forward to reading your posts!
 

 

Best regards,

 

Dimitris

C2C EMEA Community Manager

 

 

Userlevel 1

Hey @ChambiarGirl ,

Google Cloud Platform does offer speech-to-text-transcription and natural language processing with API support for sentiment analysis. While it provides insights into conversation, identifying specific nuances like stutters or unpreparedness might require customization. It generates reports and supports data visualization, but for detailed conversation analytics, additional development or specialized platforms may be needed.

Hope this helps. Let me know 😊

Reply