让大模型在知识图谱中做层次推理，提升问答准确性。
HG-RAG: Hierarchy-Guided Retrieval-Augmented Generation for Structured Knowledge Graphs
- 基于层级图谱遍历，动态扩展查询上下文
- 在多跳、关系推理任务上显著优于传统检索
- 适合需要结构化知识的复杂问答场景
检索增强生成（RAG）能有效提升大语言模型在宽泛上下文下的输出质量。然而，现有RAG系统通常从扁平文档库中检索，难以应对涉及层级或关系推理的结构化知识查询。本文提出HG-RAG（Hierarchy-Guided RAG），通过在分层知识图谱上进行图遍历，向语言模型提供结构化上下文。其检索流程首先定位查询中的命名实体锚点，随后向上追溯父节点、横向扩展关联邻居、必要时向下遍历子节点以丰富上下文。在包含18至800个节点的三个世界规模下，针对局部事实、层级、邻域和多跳四类查询进行评估。结果表明，HG-RAG在层级、关系及多跳推理任务中持续优于扁平检索基线，同时降低幻觉率并保持局部一致性。
原文摘要 · Abstract (English)
Retrieval Augmented Generation (RAG) has proven to be a widely successful process at improving the quality of outputs from a Large Language Model (LLM) for wider context. However, RAG systems typically retrieve context from flat document stores, which struggles when queries require hierarchical or relational reasoning across structured knowledge. I present HG-RAG (Hierarchy-Guided RAG), a framework that performs graph-traversal over a hierarchical knowledge graph to deliver structured context to a language model. My retrieval pipeline resolves a named entity anchor from the query, then expands context upward through parent nodes, laterally through relational neighbors, and downward through child nodes when needed. I evaluate HG-RAG against a dense retrieval baseline across three world scales (18-800 nodes) with four query types: local fact, hierarchical, neighborhood, and multi-hop. Results show HG-RAG consistently outperforms the flat baseline on hierarchical, relational, and multi-hop reasoning tasks, while reducing hallucination and maintaining locality coherence.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。