I want to build a chatbot based on Neo4j graph database, to perform query disambiguation. My queries would be related to tech products like outlook, webex. And solution would be knowledge articles that can help resolve the queries.
So I am thinking of starting with an entity extractor, that can retrieve product (outlook/webex) subentity (email) and issue (Error in sending emails)
If in case there is a missing entity, I want to check it from the user by giving him interactive options.
So the arch would be like
(Product) --> (Entity) -[:type]-> (Issue)
Can you suggest if this a proper data model,
Also the chat platform can be dynamic, how to find the closest node or relationship based on similar entity in the user query e.g. (add can be similar to create)