Hey there!
Rasa NLU seems to make it reasonably easy to train models to classify text and entities.
I have a similar use case right now where I want to classifiy a sentence and to draw out if it falls into one of two categories and learn some entities on the way. However, these sentences are not conversational inquiries like you would have them from people speaking to chatbots but they are reviews of EV charging stations. Some sentences are quite long and complex.
Has anybody heard of people using rasa for something else than ChatBot conversations?
I’ve tried to train the model with two intents for which I had 20-30 examples each but my results weren’t very good so far because the classifier often classifies quite random sentences very confidently (90%-97%) as one of the two intents.