ContextRAG: 无提取的分层图构建用于检索增强生成
ContextRAG: Extraction-Free Hierarchical Graph Construction for Retrieval-Augmented Generation
AI总结：
本文提出ContextRAG，一种无需大型语言模型提取实体和关系的图检索增强生成系统，通过残差量化k均值和Formal Concept Analysis方法构建模糊概念图，在130个任务的UltraDomain子集中实现了33.6%的F1分数，显著优于传统方法。
摘要：
图结构的检索增强生成（RAG）系统能够提高多跳问题的答案质量，但许多现有系统依赖大型语言模型（LLMs）在索引过程中提取实体、关系和摘要。这些调用会增加随语料库大小增长的token和时间成本。我们提出了ContextRAG，一种图RAG系统，其图拓扑结构无需LLM进行实体或关系提取。ContextRAG通过残差量化k均值和带有Lukasiewicz残余逻辑的Formal Concept Analysis，在片段嵌入上构建模糊概念图。通过软模糊连接和meet操作诱导桥状和meet衍生的上下文节点，而非LLM生成的图边。在130个任务的UltraDomain子集中，ContextRAG用30次LLM调用和22,073个token构建其索引。相比之下，一个本地HiRAG再现压力测试在20个任务子集上需要870次索引调用和3.54M个token才能在图构建过程中失败；线性外推到130个任务意味着超过23M个索引token。ContextRAG在整体上获得33.6%的F1分数，在多跳任务上获得36.8%的F1分数。激活分析显示，检索到至少一个由lattice衍生节点的前五查询在F1上比未检索到的查询高出+3.9个百分点；这种关联是诊断而非因果的。
英文摘要：
Graph-structured retrieval-augmented generation (RAG) systems can improve answer quality on multi-hop questions, but many current systems rely on large language models (LLMs) to extract entities, relations, and summaries during indexing. These calls add token and wall-clock costs that grow with corpus size. We present ContextRAG, a graph RAG system whose graph topology is constructed without LLM-based entity or relation extraction. ContextRAG derives a fuzzy concept graph over chunk embeddings using residual-quantization k-means and Formal Concept Analysis with Lukasiewicz residuated logic. Bridge-like and meet-derived context nodes are induced by soft fuzzy join and meet operations, rather than by LLM-written graph edges. On a 130-task UltraDomain subset, ContextRAG builds its index with 30 LLM calls and 22,073 tokens. In contrast, a local HiRAG reproduction stress test required 870 indexing calls and 3.54M tokens on a 20-task subset before failing during graph construction; linear extrapolation to 130 tasks implies over 23M indexing tokens. ContextRAG obtains 33.6% F1 overall and 36.8% F1 on multi-hop tasks. An activation analysis shows that queries retrieving at least one lattice-derived node in the top five achieve +3.9 percentage points F1 over queries that do not; this association is diagnostic rather than causal.
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。
