Building a chatbot for legal questions: exploring complex Relationships between data

Hello guys, I’m working on a chatbot that answers questions about a specific area of legislation. The available documents include legal texts and other related materials. I’m exploring possible ways to represent the complexity of relationships between various texts.

So far, I’ve found: – Llama index’s indexing (especially the keyword tablet index) – Building a relational DB

In your experience, do you have any other possible methodologies? Thanks!