Data Logging and Enhanced Speech Models

Google works continuously to improve its products, including the use of customer data to refine those products. To help Dialogflow to better suit your needs, you can opt into the Dialogflow Data Logging program. The program allows Google to improve the quality of Dialogflow by using customer data for speech recognition improvements, model training, testing, etc.

When you opt-in, you receive access to a special set of machine learning models for speech recognition. Google creates and improves these enhanced models based upon data collected through data logging. For example, if you send a request specifying a phone call model, Dialogflow uses an enhanced phone call model to process the request. Using the enhanced model, Dialogflow can more accurately recognize speech captured from a phone call and therefore produce a more accurate transcription of the audio data.

Data Privacy and Security

When you opt into the program, Google does not log all of your data: Google only uses the data sent to Dialogflow on the project with data logging enabled. You can continue to use other Google Cloud projects without data logging enabled, sending your transcription requests to Dialogflow for those projects with Google using your data only as needed to provide the service.

You continue to maintain full ownership of all data that you upload to a project with data logging enabled. However, Google retains ownership of the actual models trained from that data. For more details about the terms and conditions for data logging, see the terms for data logging.

Only a limited set of authorized Google employees have access to your data collected by data logging. Google uses this data solely to train and improve Google products and services. Google does not use your data or any personally identifiable information contained in your collected data to target products, services, or advertising towards you, your users, or your customers.

For more information about Google's security policies, see the Google Cloud security whitepaper.

Enabling data logging and using enhanced models

To enable data logging:

  1. Go to the Dialogflow Console
  2. Select your agent
  3. Click the gear icon settings next to the agent name to edit its settings
  4. Select the Speech tab
  5. Scroll to the Enhanced Speech Recognition Models section
  6. Toggle Enable data logging and enhanced models on

To use an enhanced model, you just need to specify a speech model in your API request. If you do not specify a speech model, Dialogflow does not use an enhanced model.

Disabling data logging

You can disable data logging for your project at any time. Review the terms for data logging to read about what happens to your data when you disable data logging for a project.

To disable data logging:

  1. Go to the Dialogflow Console
  2. Select your agent
  3. Click the gear icon settings next to the agent name to edit its settings
  4. Select the Speech tab
  5. Scroll to the Enhanced Speech Recognition Models section
  6. Toggle Enable data logging and enhanced models off
Hai trovato utile questa pagina? Facci sapere cosa ne pensi:

Invia feedback per...

Dialogflow Documentation
Hai bisogno di assistenza? Visita la nostra pagina di assistenza.