Azure Bot using Language Service what's the preferred method

Question

I'd like to setup a chat bot that will both answer Q&A and perform some custom actions based on intent and entities. All tutorials reference QnA Maker and LUIS which are deprecated and now Language Service's Conversational language understanding and Custom question answering features are recommended to use. But how to use both of them in one bot? I managed to run Custom question answering using Bot Framework Composer tool and changing Recognizer to custom one. But how can I now use Conversational language understanding to extract intents and entities and perform actions based on it?

Answer 1

hope the following idea helps you

• Set up a Conversational Language Understanding project, train your model, and deploy it as a REST API (ensure you have the endpoint.
• Create a new Bot Framework Composer project.
• Configure the "Recognizer" to use your Custom Question Answering project.
  • in your main dialog add HTTP request action and send a request to CLU
  • save the answer in a variable like dialog.cluResult
  • If it's a question-related intent, send it to your Custom Question Answering component.
  • If it's another intent, trigger the appropriate dialog within Composer to handle the custom action.