LEGO-GraphRAG: Modularizing Graph-based Retrieval-Augmented Generation for Design Space Exploration

Yukun Cao, Zengyi Gao, Zhiyang Li, Xike Xie, S. Kevin Zhou +1 more
11/6/2024
cs.AIcs.CLcs.DB

No official code found. Best implementation: microsoft/graphrag (34,724 stars, MIT, last commit 2026-04-30). 1 more below.

Abstract

GraphRAG integrates (knowledge) graphs with large language models (LLMs) to improve reasoning accuracy and contextual relevance. Despite its promising applications and strong relevance to multiple research communities, such as databases and natural language processing, GraphRAG currently lacks modular workflow analysis, systematic solution frameworks, and insightful empirical studies. To bridge these gaps, we propose LEGO-GraphRAG, a modular framework that enables: 1) fine-grained decomposition of the GraphRAG workflow, 2) systematic classification of existing techniques and implemented GraphRAG instances, and 3) creation of new GraphRAG instances. Our framework facilitates comprehensive empirical studies of GraphRAG on large-scale real-world graphs and diverse query sets, revealing insights into balancing reasoning quality, runtime efficiency, and token or GPU cost, that are essential for building advanced GraphRAG systems.

View on arXivView PDF

Code Implementations(2)

microsoft/graphragCommunity76%
MIT

A modular graph-based Retrieval-Augmented Generation (RAG) system

34,7243,657HTML, JinjaMar 27, 20245 months agoMIT
gptgpt-4gpt4graphragllm+2 more
gzy02/LEGO-GraphRAGCommunity56%
MIT

Source code and supplimentary materials for "LEGO-GraphRAG: Modularizing Graph-based Retrieval-Augmented Generation for Design Space Exploration"

122C, C++Jan 1, 20251 years agoMIT

Similar Papers

When to use Graphs in RAG: A Comprehensive Analysis for Graph Retrieval-Augmented Generation

90% match1 repo4492025

Graph Retrieval-Augmented Generation: A Survey

89% match1 repo2852024

A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models

89% match1 repo2.7k2025

Use Graph When It Needs: Efficiently and Adaptively Integrating Retrieval-Augmented Generation with Graphs

88% match2026

Weak-to-Strong GraphRAG: Aligning Weak Retrievers with Large Language Models for Graph-based Retrieval Augmented Generation

86% match2025

LinearRAG: Linear Graph Retrieval Augmented Generation on Large-scale Corpora

86% match1 repo5582025

Cite this paper

@article{cao2024legographrag,
  title  = {LEGO-GraphRAG: Modularizing Graph-based Retrieval-Augmented Generation for Design Space Exploration},
  author = {Yukun Cao and Zengyi Gao and Zhiyang Li and Xike Xie and S. Kevin Zhou and Jianliang Xu},
  year   = {2024},
  eprint = {2411.05844},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2411.05844v3}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2411.05844.svg)](https://codeofpaper.com/paper/2411.05844)

Discussion