Hi, Please guide me to detect question and answers from pdf documents containing medicinal drug related information to create knowledge base for chatbot as I am completely new in this.
Chatbot contains three parts:
- NLU - natural language understandings (which is similar to text classification) identifies the need of the user input.
- NLG - Which provides the response based on the NLU outcome
- State machine: in between NLU / NLG to maintain the conversation flow.
By combining this three parts, could develop the chatbot for any specific purpose.