Retrieval Augmented Generation (RAG) on large technical reports
Retrieval Augmented Generation (RAG) is seeing rapid adoption in industry to enable employees to query information captured in proprietary data for their organisation.
Henri Scaffidi’s work undertaken as part of a MRIWA funded Co-operative Education for Enterprise Development (CEED) has now been published in the Journal Transactions on Graph Data and Knowledge (TGDK).
His paper, GraphRAG on Technical Documents – Impact of Knowledge Graph Schema, identified there is an optimal level of abstraction when defining entity types for GraphRAG pipelines. A schema that is too specific may constrain the large language models and reduce recall, while one that is too generic may overlook important distinctions.
Striking the right balance appears to be a critical consideration in knowledge graph schema design for GraphRAG.
These results suggest that a domain-specific schema can assist GraphRAG with entity extraction on technical texts, leading to better response quality.
Academic supervisors included Prof. Melinda Hodkiewicz and Dr Caitlin Woods.
The journal article can be accessed here: https://doi.org/10.4230/TGDK.3.2.3
Page was last reviewed 21 October 2025