2026-11-12 · America/Los_Angeles · 社区动态 · #11

Toy project with ontologies + an LLM agent. Which stack would you use?

Hi all, I’m putting together a toy project for an exam to see how well ontologies and an LLM agent work together. The domain will probably be insurance or banking (haven’t decided yet, I was thinking about the underwriting process). The idea is to load document-extracted data as RDF, back it with an ontology, and let an agent answer questions through a few tools plus some read-only queries. What I’m hoping to get out of the ontology: fewer tokens, because the agent works from a compact schema and makes fewer exploratory calls better guidance on what it can ask next, from domain/range and the…

热度 44.2 / 100；排名与评分保留该期记录。