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refines LLM outputs, effectively mitigating issues such as ``hallucination'', lack of domain-specific knowledge, and outdated information.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 paper 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.
Paper Link: https://arxiv.org/abs/2408.08921
- [2024/9/10] We released the second version and created the repository on GitHub.
- [2024/8/15] We released the first version of our survey on arXiv.
- Overview of GraphRAG
- Graph-Based Indexing
- Graph-Guided Retrieval
- Graph-Enhanced Generation
- Downstream Tasks
- Citation
- Contact Us
We divide GraphRAG into three stages: G-Indexing, G-Retrieval, and G-Generation. We categorize the retrieval sources into open-source knowledge graphs and self-constructed graph data. Various enhancing techniques like query enhancement and knowledge enhancement may be adopted to boost the relevance of the results. Unlike RAG, which uses retrieved text directly for generation, GraphRAG requires converting the retrieved graph information into patterns acceptable to generators to enhance the task performance.
The construction and indexing of graph databases form the foundation of GraphRAG, where the quality of the graph database directly impacts GraphRAG's performance.
In GraphRAG, the retrieval process is crucial for ensuring the quality and relevance of generated outputs by extracting pertinent and high-quality graph data from external graph databases. However, retrieving graph data presents two significant challenges: (1) Explosive Candidate Subgraphs: As the graph size increases, the number of candidate subgraphs grows exponentially, requiring heuristic search algorithms to efficiently explore and retrieve relevant subgraphs. (2) Insufficient Similarity Measurement: Accurately measuring similarity between textual queries and graph data necessitates the development of algorithms capable of understanding both textual and structural information. Considerable efforts have previously been dedicated to optimizing the retrieval process to address the above challenges. This survey focuses on examining various aspects of the retrieval process within GraphRAG, including the selection of the retriever, retrieval paradigm, retrieval granularity, and effective enhancement techniques.
The generation stage is another crucial step in GraphRAG, aimed at integrating the retrieved graph data with the query to enhance response quality. In this stage, suitable generation models must be selected based on the downstream tasks. The retrieved graph data is then transformed into formats compatible with the generators. The generator takes both the query and the transformed graph data as inputs to produce the final response. Beyond these fundamental processes, generative enhancement techniques can further improve the output by intensifying the interaction between the query and the graph data and enriching the content generation itself.
Will be updated soon!
If you find this survey useful for your research or development, please cite our paper:
@misc{peng2024graphragsurvey,
title={Graph Retrieval-Augmented Generation: A Survey},
author={Boci Peng and Yun Zhu and Yongchao Liu and Xiaohe Bo and Haizhou Shi and Chuntao Hong and Yan Zhang and Siliang Tang},
year={2024},
eprint={2408.08921},
archivePrefix={arXiv},
primaryClass={cs.AI},
url={https://arxiv.org/abs/2408.08921},
}
If you have any questions or suggestions, please feel free to contact us via:
Email: [email protected]