是否需要GraphRAG？从基础RAG到基于图/智能体的上下文优化解决方案
Is GraphRAG Needed? From Basic RAG to Graph-/Agentic Solutions with Context Optimization
AI总结：
本文提出一个框架，比较常规RAG、GraphRAG、Modular RAG和Agentic RAG在9种标准化场景下的性能，并引入上下文工程方法减少19%-53%的token使用，同时发现检索-生成差距，表明扩展检索未成比例提升生成质量。
摘要：
随着GraphRAG和Agentic RAG等高级RAG变体的出现，一个关键问题是何时以及如何使用它们。在此，我们引入一个框架，用于在半结构化知识库上评估和比较不同的RAG场景，包括常规RAG、GraphRAG、Modular RAG和Agentic RAG。我们提供了9种标准化RAG场景的实现，并进行了全面比较的实验。这些场景针对数据和领域限制的实际用例设计，涵盖从简单的文档检索到高级功能，如混合文本-图检索、与计算或预定义的领域知识图谱集成、智能体多步规划以及智能体-图集成。此外，我们提出了一种新颖的上下文工程方法，用于GraphRAG和Agentic RAG，解决了上下文/内存溢出问题，通过新的表示和智能体循环设计有效管理文本和图检索，使token使用量减少19%-53%。此外，进一步分析发现了一个检索-生成差距，即扩展检索并未成比例地提高生成质量，表明面向检索的指标夸大了高级检索的益处。这项工作为构建生产级智能RAG系统提供了何时以及如何使用它们的数据驱动见解。
英文摘要：
As advanced RAG variants like GraphRAG and Agentic RAG emerge, one leading question is when and how to use them. Here, we introduce a framework for different RAG scenarios evaluation and comparison on semi-structured knowledge bases, including regular RAG, GraphRAG, Modular RAG and Agentic RAG. We provide implementation for 9 standardized RAG scenarios, and conduct experiments for a comprehensive comparison. These scenarios are designed for real use cases regarding data and domain restrictions, spanning from simple document-based retrieval to advanced features such as hybrid text-graph retrieval, integration with computed or pre-defined domain knowledge graphs, agentic multi-step planning, and agent-graph integration. Besides, we present a novel context engineering method for GraphRAG and Agentic RAG, addressing the context/memory overflow issues, efficiently managing text and graph retrievals with new representations and agentic loop design, leading to 19%-53% reduction on token usage. Moreover, further analysis identifies a retrieval-generation gap where expanded retrieval does not proportionally improve generation quality, suggesting retrieval-oriented metrics overstate advanced retrieval benefits. This work provides data-driven insights on when and how to use them for building production-ready intelligent RAG systems.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
