I want to get message enter by user for Preprocess before tranfer this message to NLU model , how can i do it?
It is recommended that you use rasa as an HTTP service, so that you can use the interface of rasa flexibly.
First, preprocess message, then, use [http://localhost:5005/model/parse] to parse message, use [http://localhost:5005/conversations/{conversation_id}/messages] to change the tracker. use [http://localhost:5005/conversations/{conversation_id}/predict] to predict next action.
1 Like
Create a custom nlu component at the very beginning of your pipeline.
1 Like