Botfront 0.27 - Handling large Rasa NLU datasets in style


Botfront 0.27 - Handling large Rasa NLU datasets in style

We just released Botfront 0.27 with a new NLU experience optimized for large Rasa NLU datasets.

A new and better NLU experience

We have rewritten the whole NLU experience to support larger datasets. The visible part (shown above) is a new infinite scrolling NLU table where you can select multiple rows, change intents in batches, use keyboard shortcuts, and tag entities more easily.

Better entity tagging

We improved the tagging and added support for roles and groups in entities that Rasa recently made available:

Before roles you had to create two separate entities. Roles allow to share the features learned and avoids having to duplicate your training data.

Regex support

You can now specify regexes to use with the RegexFeaturizer.

regex table

You can install Botfront 0.27 with npm install -g botfront or contact us to try Botfront Cloud.