How to use RASA NLU with RASA CORE

Question

I am new to chatbot application and RASA as well, can anyone please help me to understand how should i use RASA NLU with RASA CORE.

Answer 1

RASA NLU is the natural language understanding piece, which is used for taking examples of natural language and translating them into "intents." For example: "yes", "yeah", "yep" and "for sure" would all be translated into the "yes" intent.

RASA CORE on the other hand is the engine that processes the flow of conversation after the intent of the user has already been determined. RASA CORE can use other natural language translators as well, so while it pairs very nicely with RASA NLU they don't both have to be used together.

As an example if you were using both:

• User says "hey there" to RASA core bot
• Rasa core bot calls RASA NLU to understand what "hey there" means
• RASA NLU translates "hey there" into intent = hello (with 85% confidence)
• Rasa core receives "hello" intent
• Rasa core runs through it's training examples to guess what it should do when it receives the "hello" intent
• Rasa core predicts (with 92% confidence) that it should respond with the "utter_hello" template
• Rasa core responds to user "Hi, I'm your friendly Rasa bot"

Hope this helps.