Skip to main content


         This documentation site is for previous versions. Visit our new documentation site for current releases.      
 

Best practices for cleaning up training data in an IVA

Updated on July 22, 2021

Before you apply changes to the text analytics model for Pega Intelligent Virtual Assistant™ (IVA), ensure that you correct each training data record by fixing issues in the text and removing any misplaced characters in the content. Correcting training data ensures that the IVA learns only from properly formatted sample data. Eliminating mistakes when you train the IVA helps to improve the accuracy of the model.

To clean up each training record for an IVA, ensure that you remove trailing white spaces, non-alphanumeric characters, and typos. You can also eliminate incomplete tags, missing characters, and misspelled words. For example, you can remove from a data record such characters as # or misplaced apostrophes and quotation marks in sentences and phrases. To learn more about editing a training data record, see Correcting training data in an IVA.

  • Previous topic Best practices when building rule-based entities in an IVA
  • Next topic Troubleshooting the conversational channel

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us