Using graph databases for chatbots

Graph databases can be used model the relationships between entities, e.g what products are available in a store. Another example would be a Digital Twin, which is the virtual representation of things in the real world.

Digital twins can then be used to detect dependencies and simulate the behavior of solutions before they are created. The knowledge that is contained in a Digital Twin can then be used to

  • dynamically control the flow of the conversation with chatbots,
  • transfer the knowledge captured in the graph between different chatbot solutions, or
  • bridge previously separate semantic domains (e.g. research and production of a manufacturer) by combining different knowledge graphs.

This repo contains a simple technical integration between a chatbot based on the Watson Assistant service and a graph database solution using a Retail store scenario.

Leave a Reply