Graph Retrieval-Augmented Generation: A Survey
Генерация с дополнением извлечением на основе графов: обзор
2025-11-19
SCID: 54.1/k5m5385r
Discuss with AI
GraphRAGgraph-based indexinggraph-enhanced generationgraph-guided retrievalretrieval-augmented generation
Figures from the paper
Abstract (AI)
Recently, Retrieval-Augmented Generation (RAG) has achieved remarkable success in addressing the challenges of Large Language Models (LLMs) without necessitating retraining. By referencing an external knowledge base, RAG refines LLM outputs, effectively mitigating issues such as “hallucination,” lack of domain-specific knowledge, and outdated information. However, the complex structure of relationships among different entities in databases presents challenges for RAG systems. In response, GraphRAG leverages structural information across entities to enable more precise and comprehensive retrieval, capturing relational knowledge and facilitating more accurate, context-aware responses. Given the novelty and potential of GraphRAG, a systematic review of current technologies is imperative. This article provides the first comprehensive overview of GraphRAG methodologies. We formalize the GraphRAG workflow, encompassing Graph-Based Indexing, Graph-Guided Retrieval, and Graph-Enhanced Generation. We then outline the core technologies and training methods at each stage. Additionally, we examine downstream tasks, application domains, evaluation methodologies, and industrial use cases of GraphRAG. Finally, we explore future research directions to inspire further inquiries and advance progress in the field. In order to track recent progress, we set up a repository at https://github.com/pengboci/GraphRAG-Survey .
Key Findings
1
GraphRAG uses relational and structural information among entities to enable more precise, comprehensive, and context-aware retrieval than conventional RAG.
2
The authors identify future research directions and provide a public repository for tracking developments in GraphRAG.
3
The review summarizes core technologies and training methods across all three GraphRAG stages.
4
The survey analyzes GraphRAG downstream tasks, application domains, evaluation methodologies, and industrial use cases.
5
The survey presents the first comprehensive review of GraphRAG methodologies and formalizes its workflow into graph-based indexing, graph-guided retrieval, and graph-enhanced generation.
Research Object
Graph Retrieval-Augmented Generation (GraphRAG) systems
Research Subject
GraphRAG methodologies, workflow stages, technologies, training methods, applications, evaluation, and future research directions
Publication Details
Publication Date
2025-11-19
Journal
Publisher
ISSN
Cited by
422
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai9
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer2019
Dense Passage Retrieval for Open-Domain Question Answering2020
Retrieval-Augmented Generation for Large Language Models: A Survey2023
Lost in the Middle: How Language Models Use Long Contexts2024
Unifying Large Language Models and Knowledge Graphs: A Roadmap2024
GPT-4 Technical Report2023
Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering2021
Retrieval-Augmented Generation for AI-Generated Content: A Survey2026
LightRAG: Simple and Fast Retrieval-Augmented Generation2025