Let's say I have a closed-domain chatbot, and it's knowledge base is in Finance. Thus, of course I want the chatbot to answer questions that the user might have like, "How is the best way to save money?", or "What are my spending habits like the past 3 months?". However, I would also want the chatbot to be able to remember the conversation (up to a certain date) as well as model what generative models like BlenderBot (Facebook AI Research) or Meena (Google AI) are able to do. The ability to respond to anyway the conversation flows?
Is it a question of just building out my intents to classify? Or should I be looking to involve both intent classification and seq2seq modeling? If so, how would one do that? Any blogs, tutorials would be helpful! Thanks!