How to make training data with 1000s of features

Hi there! I am working on developing a chatbot using RASA. The bot needs to fetch data from API. But there are 100s and 1000s of features which bot should be able to understand so that it can get the required data back to user. So how to train the bot with data consisting lots of feature_type(entity) values.

PS: I tried mentioning all the features in the lookup table but it isn’t working.

Thanks!! Your help will be appreciated. @Juste

Can you use regular expressions to separate the entities, and use the API to validate the values?

But those entities are nothing but the normal English words. How to use regular expression to distinguish it

Can you post a few examples of values, to understand better what you mean?

  • Show me [Bios Version] (feature_type)
  • Show me [Chassis Model] (feature_type)
  • Can you please tell the [IP Address] (feature_type) ?
  • Can you please tell the [DNS Rac name] (feature_type)?

So here I’ve 1000 such feature types(entity value). How to train bot to identify and extract the correct entities? How can I use regular expression for this?