3 Recognizing Intent from the User Query

 

This chapter covers

  • Choosing your approach for intent recognition
  • Understanding when to apply NLP
  • Using Luis for implementing NLP
  • Using Intent Recognition in our health assistant bot

In the last chapter, we learnt about Conversational Design Patterns and in the chapter before that we built a basic echo chatbot. Our bot currently doesn’t really do anything other than simply echoing back the message. We will use the principles we learnt about conversational design inside the echo chatbot and make it more functional.

3.1   What is intent?

3.1.1   Designing our Intents

3.2   Intent Recognition with simple string matching

3.2.1   Implementing string matching in our chatbot

3.2.2   Problems with simple string matching

3.3   Intent recognition with AIML

3.3.1   Creating a chatbot with AIML

3.4   Intent recognition with NLP

3.4.1   What is NLP?

3.4.2   Natural Language Understanding (NLU)

3.4.3   Natural Language Generation (NLG)

3.5   A deeper look into Intent Recognition using NLP

3.6   LUIS overview

3.6.1   Creating a LUIS application

3.6.2   Adding intents

3.7   Integrating Luis into our bot

3.7.1   Testing the endpoint

3.7.2   How does Luis pass intents?

3.7.3   Adding Luis recognizer to our code

3.7.4   Testing our bot

3.8   Alternatives to Luis

3.8.1   Dialogflow

3.8.2   wit.ai

3.8.3   AWS Lex and IBM Watson

3.8.4   rasa.ai

3.9   Summary