Correct a False Positive (Disagree with the virtual assistant’s prediction of intent)

Updated on December 12, 2022

For when you do not agree with the virtual assistant’s automatically predicted intent, you can take action to classify the question under a different existing intent.

How is this useful?
Some intents might not be correctly predicted by the virtual assistant since it has yet to be trained, so classifying new examples as they come in can help improve relevance of the virtual assistant’s response.

Steps to classify the question under a different intent:

  1. If you disagree with the bot’s prediction, click on the “Classify” icon 
  2. You can then type in the correct intent in the search bar.
  3. You can enter the name of the intent or keywords into the search bar.
  4. In this example, we typed “escalation”, and the matching intent showed up below. Click on it to select.
  5. Click on the tick that appeared on the right to confirm.
  6. You can then see the new question in the ‘Training Dataset’ tab when you select the intent you had chosen in the above steps .
  7. When you are done, always click ‘Save’, ‘Train’ and ‘Publish’ to publish the changes.
Did we help you?

Learn more about KeyReply

To see our Conversational AI in action, sign-up for a guided walk-through and see how our platform can be tailored for your organization’s needs