arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2510.09156cs.LG
收藏

Agentic-KGR:通过多智能体强化学习实现共演化的知识图谱构建

Agentic-KGR: Co-evolutionary Knowledge Graph Construction through Multi-Agent Reinforcement Learning

Huawei Technologies Co., Ltd.(华为技术有限公司)

Jing Li, Zhijie Sun, Zhicheng Zhou, Suming Qiu, Junjie Huang, Haijia Sun, Linyuan Qiu

AI总结:

针对静态知识库覆盖不足和时效性老化问题,提出Agentic-KGR框架,通过多轮强化学习实现LLMs与知识图谱共演化,引入动态模式扩展、检索增强记忆和多尺度提示压缩,显著提升知识抽取及下游QA任务的准确率与覆盖率。

摘要:

当前的知识增强大型语言模型(LLMs)依赖于静态的、预先构建的知识库,这些知识库存在覆盖范围不足和时效性老化的问题,限制了它们在动态信息环境中的有效性。我们提出了Agentic-KGR,这是一个通过多轮强化学习(RL)实现LLMs与知识图谱(KGs)共演化的新框架。我们的方法引入了三项关键创新:(1)动态模式扩展机制,在训练过程中系统地扩展图本体,突破预定义的边界;(2)检索增强记忆系统,通过持续优化实现模型参数与知识结构的协同共演化;(3)可学习的多尺度提示压缩方法,通过自适应序列优化在降低计算复杂度的同时保留关键信息。实验结果表明,在知识抽取任务中,该方法相比监督基线和单轮RL方法取得了显著提升。当与GraphRAG集成时,我们的方法在下游QA任务中取得了优越的性能,与现有方法相比,在准确率和知识覆盖率方面均有显著提升。

英文摘要:

Current knowledge-enhanced large language models (LLMs) rely on static, pre-constructed knowledge bases that suffer from coverage gaps and temporal obsolescence, limiting their effectiveness in dynamic information environments. We present Agentic-KGR, a novel framework enabling co-evolution between LLMs and knowledge graphs (KGs) through multi-round reinforcement learning (RL). Our approach introduces three key innovations: (1) a dynamic schema expansion mechanism that systematically extends graph ontologies beyond pre-defined boundaries during training; (2) a retrieval-augmented memory system enabling synergistic co-evolution between model parameters and knowledge structures through continuous optimization; (3) a learnable multi-scale prompt compression approach that preserves critical information while reducing computational complexity through adaptive sequence optimization. Experimental results demonstrate substantial improvements over supervised baselines and single-round RL approaches in knowledge extraction tasks. When integrated with GraphRAG, our method achieves superior performance in downstream QA tasks, with significant gains in both accuracy and knowledge coverage compared to existing methods.

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。

↑