arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2504.06271cs.IRcs.AIcs.CL
收藏

ER-RAG:利用基于ER的异构数据源统一建模增强RAG

ER-RAG: Enhance RAG with ER-Based Unified Modeling of Heterogeneous Data Sources

Peking University(北京大学);Huawei Technologies Co., Ltd(华为技术有限公司)

Yikuan Xia, Jiazun Chen, Yirui Zhan, Suifeng Zhao, Weipeng Jiang, Chaorui Zhang, Wei Han, Bo Bai, Jun Gao

AI总结:

ER-RAG提出基于实体-关系模型的统一框架,通过ER API和两阶段生成过程整合异构数据源,在2024 KDDCup CRAG挑战赛中全赛道获胜,性能超越混合方法并显著加速检索。

摘要:

大型语言模型(LLMs)在问答(QA)任务中表现出色,而检索增强生成(RAG)通过纳入来自网页、数据库和知识图谱等多种来源的外部证据来提升其精确性。然而,当前的RAG方法依赖针对单个数据源的智能体特定策略,这在低资源或黑盒环境中带来挑战,并且当证据分散在多个来源时会使操作复杂化。为解决这些局限性,我们提出ER-RAG,一个利用实体-关系(ER)模型统一异构数据源证据集成的框架。ER-RAG通过基于ER的API,以GET和JOIN操作标准化实体检索和关系查询。它采用两阶段生成过程:首先,偏好优化模块选择最优来源;其次,另一个模块根据来源模式构建API链。这种统一方法支持高效微调并实现跨多种数据源的无缝集成。ER-RAG在2024年KDDCup CRAG挑战赛的全部三个赛道中获胜,证明了其有效性,在使用8B LLM骨干网络时取得了与商业RAG流水线相当的性能。它在LLM得分上比混合竞争对手高出3.1%,并将检索速度提升了5.5倍。

英文摘要:

Large language models (LLMs) excel in question-answering (QA) tasks, and retrieval-augmented generation (RAG) enhances their precision by incorporating external evidence from diverse sources like web pages, databases, and knowledge graphs. However, current RAG methods rely on agent-specific strategies for individual data sources, posing challenges low-resource or black-box environments and complicates operations when evidence is fragmented across sources. To address these limitations, we propose ER-RAG, a framework that unifies evidence integration across heterogeneous data sources using the Entity-Relationship (ER) model. ER-RAG standardizes entity retrieval and relationship querying through ER-based APIs with GET and JOIN operations. It employs a two-stage generation process: first, a preference optimization module selects optimal sources; second, another module constructs API chains based on source schemas. This unified approach allows efficient fine-tuning and seamless integration across diverse data sources. ER-RAG demonstrated its effectiveness by winning all three tracks of the 2024 KDDCup CRAG Challenge, achieving performance on par with commercial RAG pipelines using an 8B LLM backbone. It outperformed hybrid competitors by 3.1% in LLM score and accelerated retrieval by 5.5X.

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。

↑