--- title: "A Survey of Graph Retrieval-Augmented Generation for" description: "The survey offers no new experiments. 10 references, cited by 50 papers here. General-purpose LLMs often fall short in specialized fields." authors: ["Qinggang Zhang", "Shengyuan Chen", "Yuanchen Bei", "Zheng Yuan", "Huachi Zhou", "Zijin Hong", "Hao Chen", "Yilin Xiao", "Chuang Zhou", "Junnan Dong", "Yi Chang", "Xiao Huang"] year: 2025 venue: "arXiv:2501.13958" source: "https://arxiv.org/abs/2501.13958" doi: "10.48550/arxiv.2501.13958" arxiv: "2501.13958" licence: "CC BY 4.0 (arXiv)" licence_url: "http://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/zhang-2025-survey-graph-retrieval-augmented-generation/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhang-2025-survey-graph-retrieval-augmented-generation/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/) / A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models # A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models **Authors:** Qinggang Zhang, Shengyuan Chen, Yuanchen Bei, [Zheng Yuan](https://md.research.elamur.ai/free/en/authors/zheng-yuan/), Huachi Zhou, Zijin Hong, [Hao Chen](https://md.research.elamur.ai/free/en/authors/hao-chen/), Yilin Xiao, Chuang Zhou, Junnan Dong, [Yi Chang](https://md.research.elamur.ai/free/en/authors/yi-chang/) and Xiao Huang · **Year:** 2025 · **Venue:** arXiv:2501.13958 · **Licence:** [CC BY 4.0 (arXiv)](http://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://arxiv.org/abs/2501.13958) · [PDF](https://arxiv.org/pdf/2501.13958) · [DOI 10.48550/arxiv.2501.13958](https://doi.org/10.48550/arxiv.2501.13958) · [arXiv 2501.13958](https://arxiv.org/abs/2501.13958) · [OpenAlex](https://openalex.org/W4406840433) **Citations:** 14 citations (OpenAlex, 2026-10-14) · 50 in our library, of the 141,840 papers whose bibliographies we read ## Summary General-purpose LLMs often fall short in specialized fields. Their domain knowledge is shallow, they struggle with long chains of rule-bound reasoning, and they misread terms whose meaning depends on context. Fine-tuning on domain data can introduce hallucinations and catastrophic forgetting. Classic RAG avoids retraining, but because it retrieves flat text chunks it has its own problems. It handles complex queries poorly, cannot link knowledge spread across many sources, is constrained by the LLM's context window, and scales badly. This survey, by authors from The Hong Kong Polytechnic University and Jilin University, reviews Graph-based RAG (GraphRAG) as a remedy. GraphRAG stores knowledge as graphs of entities and relations, retrieves it with structure-aware and multi-hop methods, and feeds the result into generation. The authors organize the field around knowledge organization, retrieval and integration. They also review applications, open-source projects and benchmark datasets, and they outline open problems in knowledge quality, efficiency, generalization and security. The survey offers no new experiments. Its value is a structured map of the field and a curated resource list. ## Key points - The survey sorts GraphRAG into three families: - **knowledge-based**, where the graph itself carries the knowledge; - **index-based**, where the graph serves as an index pointing back to raw text; - **hybrid**, which combines the two. - It names four weaknesses of flat-text RAG: - difficulty understanding complex queries, e.g. missing intermediate concepts B and C when linking A to D; - poor integration of knowledge from distributed sources; - LLM context windows of roughly 2K–32K tokens; - efficiency and scalability bottlenecks. - The core chapters cover: - structured knowledge organization through knowledge carrier graphs and index graphs; - graph retrieval techniques; - methods for integrating retrieved knowledge into LLMs. - The benchmarks surveyed include WebQSP, CWQ, MetaQA, GrailQA, HotpotQA, KQAPro, FACTKG and UltraDomain. The representative systems covered include Microsoft-style GraphRAG, LightRAG, ToG, RoG, G-Retriever, GNN-RAG and MedGraphRAG. - Papers, datasets and projects are collected in the GitHub repository DEEP-PolyU/Awesome-GraphRAG. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* ## Institutions - [Hong Kong Polytechnic University](https://md.research.elamur.ai/free/en/universities/hong-kong-polytechnic-university/) — University, Hong Kong, China · 34 works · 4,476 citations - [Jilin University](https://md.research.elamur.ai/free/en/universities/jilin-university/) — University, Changchun, China · 8 works · 5,204 citations 📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,085 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,888 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works · [👁️ Computer Vision and Pattern Recognition](https://md.research.elamur.ai/free/en/shelves/computer-science/computer-vision-and-pattern-recognition/) · 1,282 works · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 523 works 🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 1,344 works · [🔍 Information retrieval & search](https://md.research.elamur.ai/free/en/topics/information-retrieval/) · 146 works ## Cite Cite as: Qinggang Zhang, Shengyuan Chen, Yuanchen Bei, Zheng Yuan, Huachi Zhou, Zijin Hong, Hao Chen, Yilin Xiao, Chuang Zhou, Junnan Dong, Yi Chang and Xiao Huang. 2025. A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models. arXiv:2501.13958. The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @misc{zhang2025survey, title = {{A Survey of Graph Retrieval-Augmented Generation for Customized Large Language Models}}, author = {Qinggang Zhang, Shengyuan Chen, Yuanchen Bei, Zheng Yuan, Huachi Zhou, Zijin Hong, Hao Chen, Yilin Xiao, Chuang Zhou, Junnan Dong, Yi Chang and Xiao Huang}, year = {2025}, doi = {10.48550/arxiv.2501.13958}, eprint = {2501.13958}, archivePrefix = {arXiv}, url = {https://doi.org/10.48550/arxiv.2501.13958} } ``` - [← previous A Comprehensive Survey on Long Context Language Modeling](https://md.research.elamur.ai/free/en/papers/liu-2025-comprehensive-survey-long-context-language/) - [next → CoT-Valve: Length-Compressible Chain-of-Thought Tuning](https://md.research.elamur.ai/free/en/papers/ma-2025-cot-valve-length-compressible-chain/) - [↑ back to papers of 2025](https://md.research.elamur.ai/free/en/papers/2025/page/2/) ## References 10 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. [3] D. Khashabi, S. Min, T. Khot, and et al., “[Unifiedqa: Crossing format boundaries with a single qa system](https://md.research.elamur.ai/free/en/papers/khashabi-2020-unifiedqa-crossing-format-boundaries-single/),” in Findings of Empirical Methods in Natural Language Processing, 2020. [4] A. Chowdhery, S. Narang, J. Devlin, and et al., “Palm: Scaling language modeling with pathways,” Journal of Machin… 2. [29] T. T. Procko and O. Ochoa, “Graph retrieval-augmented generation for [large language models: A survey](https://md.research.elamur.ai/free/en/papers/minaee-2024-large-language-models-survey/),” in Conference on AI, Science, Engineering, and Technology (AIxSET), 2024. [30] H. Han, Y. Wang, H. Shomer, and et al., “Retrieval-augmented generation with graphs (graphrag),” arXiv preprint a… 3. [54] J. Wang, J. Fu, R. Wang, and et al., “Pike-rag: specialized knowledge and rationale augmented generation,” arXiv preprint arXiv:2501.11551, 2025. [55] K. Sharma, P. Kumar, and Y. Li, “[Og-rag: Ontology-grounded retrieval-augmented generation for large language models](https://md.research.elamur.ai/free/en/papers/sharma-2025-og-rag-ontology-grounded-retrieval/),” arXiv preprint arXiv:2412.… 4. [79] M. Gubanov, A. Pyayt, and A. Karolak, “Cancerkg. org-a web-scale, interactive, verifiable knowledge graph-llm hybrid for assisting with optimal cancer treatment and care,” in International Conference on Information and Knowledge Management, 2024. [80] X. He, Y. Tian, Y. Sun, and et al., “G-retr… — [G-Retriever: Retrieval-Augmented Generation for Textual Graph Understanding and Question Answering](https://md.research.elamur.ai/free/en/papers/he-2024-g-retriever-retrieval-augmented-generation/) 5. [104] R. Ye, C. Zhang, R. Wang, and et al., “Language is all a graph needs,” in Findings of Association for Computational Linguistics, 2024. [105] R. Chen, T. Zhao, A. K. JAISWAL, and et al., “LLaGA: Large language and graph assistant,” in International Conference on Machine Learning, 2024. [106] N.… [find](https://scholar.google.com/scholar?q=%5B104%5D+R%2E+Ye%2C+C%2E+Zhang%2C+R%2E+Wang%2C+and+et+al%2E%2C+%E2%80%9CLanguage+is+all+a+graph+needs%2C%E2%80%9D+in+Findings+of+Association+for+Computational+Linguistics%2C+2024%2E+%5B105%5D+R%2E+Chen%2C+T%2E+Zhao%2C+A%2E+K%2E+JAISWAL%2C+and+et+al%2E%2C+%E2%80%9CLLaGA%3A+) 6. [128] G. Angeli, M. J. J. Premkumar, and C. D. Manning, “[Leveraging linguistic structure for open domain information extraction](https://md.research.elamur.ai/free/en/papers/angeli-2015-leveraging-linguistic-structure-open-domain/),” in Association for Computational Linguistics and International Joint Conference on Natural Language Processing, 2015. [129] W. Li, H. Zhou, J. Dong, and et al., “Construc… 7. [152] W. Hamilton, Z. Ying, and J. Leskovec, “[Inductive representation learning on large graphs](https://md.research.elamur.ai/free/en/papers/hamilton-2017-inductive-representation-learning-large-graphs/),” in Advances in Neural Information Processing Systems, 2017. [153] Y. Bei, S. Zhou, Q. Tan, and et al., “Reinforcement neighborhood selection for unsupervised graph anomaly detection,” in International C… 8. [177] X. Lin, W. Wang, Y. Li, and et al., “Data-efficient fine-tuning for llm-based recommendation,” in Conference on Information Retrieval, 2024. [178] J. He, M. Rungta, D. Koleczek, and et al., “Does prompt formatting have any impact on llm performance?” arXiv preprint arXiv:2411.10541, 2024. [179… — [ChatDoctor: A Medical Chat Model Fine-Tuned on a Large Language Model Meta-AI (LLaMA) Using Medical Domain Knowledge](https://md.research.elamur.ai/free/en/papers/li-2023-chatdoctor-medical-chat-model-fine/) 9. [203] S. Xu, L. Pang, H. Shen, and et al., “Search-in-the-chain: Interactively enhancing large language models with search for knowledge-intensive tasks,” in International World Wide Web Conference, 2024. [204] S. Geng, Z. Fu, Y. Ge, and et al., “Improving personalized explanation generation through… [find](https://scholar.google.com/scholar?q=%5B203%5D+S%2E+Xu%2C+L%2E+Pang%2C+H%2E+Shen%2C+and+et+al%2E%2C+%E2%80%9CSearch%2Din%2Dthe%2Dchain%3A+Interactively+enhancing+large+language+models+with+search+for+knowledge%2Dintensive+tasks%2C%E2%80%9D+in+International+World+Wide+Web+Conference%2C+2) 10. [229] Y. Gu, S. Kase, M. Vanni, and et al., “Beyond iid: three levels of generalization for question answering on knowledge bases,” in International World Wide Web Conference, 2021. [230] Y. Lyu, Z. Li, S. Niu, and et al., “Crud-rag: A comprehensive chinese benchmark for retrieval-augmented generati… [find](https://scholar.google.com/scholar?q=%5B229%5D+Y%2E+Gu%2C+S%2E+Kase%2C+M%2E+Vanni%2C+and+et+al%2E%2C+%E2%80%9CBeyond+iid%3A+three+levels+of+generalization+for+question+answering+on+knowledge+bases%2C%E2%80%9D+in+International+World+Wide+Web+Conference%2C+2021%2E+%5B230%5D+Y%2E+Lyu%2C+Z%2E+L) *Updated 24 October 2026*