解决黑箱知识图谱中检索不准的问题，提升大模型问答准确率
BubbleRAG: Evidence-Driven Retrieval-Augmented Generation for Black-Box Knowledge Graphs
- 通过语义锚点分组与气泡扩展，自动发现潜在证据图
- 在多跳问答任务上F1和准确率均超越现有方法
- 无需训练，可直接接入现有系统，适合知识增强应用
大语言模型在知识密集型任务中常出现幻觉。基于图的检索增强生成（RAG）虽有前景，但在黑箱知识图谱（结构与模式未知）上仍面临召回与精度双重瓶颈。本文识别出三类根本问题：语义实例不确定性、结构路径不确定性及证据对比不确定性。为此，将检索任务形式化为最优信息子图检索（OISR），证明其为NP-hard和APX-hard。提出BubbleRAG——一种无训练的流水线，通过语义锚点分组、启发式气泡扩展生成候选证据图（CEGs）、复合排序与推理感知扩展，系统优化召回与精度。在多跳问答基准测试中，BubbleRAG表现领先，显著优于多个强基线，在保持即插即用特性的同时实现最佳性能。
原文摘要 · Abstract (English)
Large Language Models (LLMs) exhibit hallucinations in knowledge-intensive tasks. Graph-based retrieval augmented generation (RAG) has emerged as a promising solution, yet existing approaches suffer from fundamental recall and precision limitations when operating over black-box knowledge graphs -- graphs whose schema and structure are unknown in advance. We identify three core challenges that cause recall loss (semantic instantiation uncertainty and structural path uncertainty) and precision loss (evidential comparison uncertainty). To address these challenges, we formalize the retrieval task as the Optimal Informative Subgraph Retrieval (OISR) problem -- a variant of Group Steiner Tree -- and prove it to be NP-hard and APX-hard. We propose BubbleRAG, a training-free pipeline that systematically optimizes for both recall and precision through semantic anchor grouping, heuristic bubble expansion to discover candidate evidence graphs (CEGs), composite ranking, and reasoning-aware expansion. Experiments on multi-hop QA benchmarks demonstrate that BubbleRAG achieves state-of-the-art results, outperforming strong baselines in both F1 and accuracy while remaining plug-and-play.
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。