I am a novice in AI and I like to build a chatbot to predict diseases using patient narration as input. Initially, I simply want to train my chatbot on 1 disease only. And once this initial milestone is accomplished then I want to train it on other diseases.

I want to know whether I should build and train my model first and then move towards building a chatbot or should i create a chatbot first and then train it on a disease.

Also please justify which approach is better and why?


1 Answer 1


The former: build and train a model first, and then think about the user interface.

Effectively, a chatbot is a user interface to your model. If you run it 'off-line' on input text and it works, then you have achieved your goal without the added complexity of driving a conversation (which is harder than one might think).

Also, building an 'abstract' chatbot devoid of any content is going to be harder. What should that chatbot talk about?

From my own experience (I work in conversational AI), you might not succeed in building a decent chatbot, especially as a novice. But you might be able to train a model to identify diseases from textual input. So if you do that first, you have at least got something! Especially if that is your main reason for the project in the first place.

  • $\begingroup$ First and foremost, thank you for your reply. I am a bit confused. Are you implying that this approach here in this blog here is not appropriate: medium.com/analytics-vidhya/… and neither this approach here: upgrad.com/blog/how-to-make-chatbot-in-python/… ? Please elaborate and explain. Once again, thank you. $\endgroup$ Commented Apr 17, 2020 at 11:36
  • 1
    $\begingroup$ Those are just introductory articles to demonstrate the concept. You are trying to build a full-blown application, not just a toy demo. $\endgroup$ Commented Apr 17, 2020 at 13:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .