Rasa: Export corrected intent's to NLU training data during interactive learning

[tiwari85aman]

Rasa Core version: (0.8.1)

Python version: 3.6.3

Operating system : Linux

Issue:As we can export "stories.md" file with interactive learning, is there any way by which we can also export the correct classifed intents during the interactive learning?

e.g.
domain.yml

intents:
  - greet
  - goodbye

nlu.md

## intent:greet
- hey
- hello
- hi

Now during interactive learning:-
user message: hiii there
bot classified it wrongly let say -- "intent: goodbye"
then i correct the intent classification-- "intent:greet"

now if i restart the chatbot it again shows that misclassifed intent ("goodbye"). how to tackle such issue. is there any way that we can export like-

temp.md

## intent:greet
- hiii there

[tmbo]

That is a very good and useful idea - it is currently not implemented, only the stories are exported at the moment.

[tiwari85aman]

@tmbo actually I am working on this, will I be able to contribute to Rasa core by creating a Pull Request if I get this feature working?

[tmbo]

Thats great! Yes of course you can open a PR and we'll discuss if there are any changes that need to be made to merge it :+1:

[wochinge]

@tmbo I guess we can close this one right?

[tmbo]

yes :+1: