GraphWalker: 通过合成轨迹课程实现基于知识图谱的代理问答
GraphWalker: Agentic Knowledge Graph Question Answering via Synthetic Trajectory Curriculum
AI总结：
GraphWalker通过自动化轨迹合成和分阶段微调解决知识图谱问答中的探索与泛化问题，提升轻量强化学习性能，在CWQ和WebQSP上取得最优效果。
摘要：
基于知识图谱的代理问答（KGQA）要求代理迭代交互知识图谱，面临训练数据稀缺和推理泛化挑战。现有方法常限制代理探索：基于提示的方法缺乏自主导航训练，现有训练流程通常限制推理到预定义轨迹。为此，本文提出GraphWalker，一种新的KGQA框架，通过自动化轨迹合成和分阶段微调解决这些挑战。GraphWalker采用两阶段SFT训练范式：首先，代理在由受限随机游走路径合成的结构多样轨迹上训练，建立对KG的广泛探索先验；其次，代理在少量专家轨迹上进一步微调，发展反思和错误恢复能力。大量实验表明，我们的分阶段SFT范式解锁了轻量强化学习（RL）阶段更高的性能上限，使GraphWalker在CWQ和WebQSP上取得最优性能。此外，GraphWalker在GrailQA和我们构建的GraphWalkerBench上的额外结果证实，GraphWalker增强了对分布外推理路径的泛化能力。代码已公开在https://github.com/XuShuwenn/GraphWalker
英文摘要：
Agentic knowledge graph question answering (KGQA) requires an agent to iteratively interact with knowledge graphs (KGs), posing challenges in both training data scarcity and reasoning generalization. Specifically, existing approaches often restrict agent exploration: prompting-based methods lack autonomous navigation training, while current training pipelines usually confine reasoning to predefined trajectories. To this end, this paper proposes GraphWalker, a novel agentic KGQA framework that addresses these challenges through Automated Trajectory Synthesis and Stage-wise Fine-tuning. GraphWalker adopts a two-stage SFT training paradigm: First, the agent is trained on structurally diverse trajectories synthesized from constrained random-walk paths, establishing a broad exploration prior over the KG; Second, the agent is further fine-tuned on a small set of expert trajectories to develop reflection and error recovery capabilities. Extensive experiments demonstrate that our stage-wise SFT paradigm unlocks a higher performance ceiling for a lightweight reinforcement learning (RL) stage, enabling GraphWalker to achieve state-of-the-art performance on CWQ and WebQSP. Additional results on GrailQA and our constructed GraphWalkerBench confirm that GraphWalker enhances generalization to out-of-distribution reasoning paths. The code is publicly available at https://github.com/XuShuwenn/GraphWalker.
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。
