Hello Team,
Just to elaborate a bit more -
Say, I have given 10 questions in my training data, with project_spacy , [10 such different equipment names in Error intent] as entities for the same intent of ‘error’. I expect rasa_nlu to understand when a user calls for of equiptment1 or any Equipment_X' which is not explicitly mentioned in my training data.
First question - do you already know all of the possible errors someone might mention? If so, you can use a simple lookup table
If the user can say many different things, then you will have to train a model to recognise these. You will have to add NLU training data, see for example this docs page