In this work, we investigate the use of Large Language Models (LLMs) within a Graph-based Retrieval Augmented Generation (RAG) architecture for Energy Efficiency (EE) Question Answering.
First, the system automatically extracts a Knowledge Graph (KG) from guidance and regulatory documents in the energy field. Then, the generated graph is navigated and reasoned upon to provide users with accurate answers in multiple languages. We incorporate these architecture into ENERGENIUS, an European project on EE.
We implement a human-based validation using the RAGAs framework properties, on a dataset that comprises a collection of documents collected from web sites in Italian and a collection of 101 questions and answers pairs about energy consumption, EE, regulations, and incentives extracted from the websites. The validation questions include:
– 25 questions & answers focusing on Italian regulation and incentives on EE;
– 25 questions & answers addressing Swiss regulations and incentives on EE;
– 51 questions & answers providing recommendations and suggestions on EE, applicable to both Italy and Switzerland.
Results confirm the potential of this architecture and identify its strengths and weaknesses. Validation results show how the system correctly answers in about three out of four of the cases (75.2%), with higher results on questions related to more general EE answers (up to 81.0%), and featuring promising multilingual abilities ($4.4% accuracy loss due to translation).
The slides of the presentation are available here:
https://www.slideshare.net/slideshow/a-graphrag-approach-for-energy-efficiency-q-a/281205134
The paper is published in the proceedings of the ICWE 2025 conference, held in Delft in July 2025.
