Simple is Effective: The Roles of Graphs and Large Language Models in Knowledge-Graph-Based Retrieval-Augmented Generation

Mufei Li, Siqi Miao, Pan Li

Keywords: Knowledge Graphs, Large Language Models, Retrieval-Augmented Generation, Retrieval

Abstract: Large Language Models (LLMs) demonstrate strong reasoning abilities but face limitations such as hallucinations and outdated knowledge. Knowledge Graph (KG)-based Retrieval-Augmented Generation (RAG) addresses these issues by grounding LLM outputs in KGs. However, current KG-based RAG frameworks struggle to optimize the trade-off between retrieval effectiveness and efficiency. We introduce SubgraphRAG, a new KG-based RAG framework. It integrates a lightweight MLP with a parallel triple-scoring mechanism for efficient and flexible subgraph retrieval while encoding directional structural distances to enhance retrieval effectiveness. The size of retrieved subgraphs can be flexibly adjusted to match the query's needs and the downstream LLM's capabilities. This design strikes a balance between model complexity and reasoning power, enabling scalable and generalizable retrieval processes. Notably, based on our retrieved subgraphs, smaller LLMs like Llama3.1-8B-Instruct deliver competitive results, while larger models like GPT-4o achieve state-of-the-art accuracy — all without fine-tuning. Extensive evaluations on the WebQSP and CWQ benchmarks highlight SubgraphRAG's strengths in efficiency, accuracy, and knowledge grounding.

Submission Number: 38