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 2512.15219cs.CLcs.AI
收藏

RFKG-CoT: 基于关系的自适应步数选择与少样本路径引导用于知识感知问答

RFKG-CoT: Relation-Driven Adaptive Hop-count Selection and Few-Shot Path Guidance for Knowledge-Aware QA

Chao Zhang, Minghan Li, Tianrui Lv, Guodong Zhou(张超,李明翰,吕天睿,周国栋)

Chao Zhang, Minghan Li, Tianrui Lv, Guodong Zhou

AI总结:

RFKG-CoT通过关系驱动的自适应步数选择和少样本路径引导,提升知识感知问答的准确性与可靠性。

摘要:

大型语言模型(LLMs)在知识密集型问答中常常由于参数知识限制而生成幻觉。尽管现有方法如KG-CoT通过整合知识图谱(KG)路径来提高可靠性,但它们存在步数选择僵化(仅由问题驱动)和推理路径利用不足(缺乏引导)的问题。为了解决这一问题,我们提出了RFKG-CoT:首先,它用关系驱动的自适应步数选择器替代了僵化的步数选择器,通过激活KG关系动态调整推理步骤(例如,1步用于直接

英文摘要:

Large language models (LLMs) often generate hallucinations in knowledge-intensive QA due to parametric knowledge limitations. While existing methods like KG-CoT improve reliability by integrating knowledge graph (KG) paths, they suffer from rigid hop-count selection (solely question-driven) and underutilization of reasoning paths (lack of guidance). To address this, we propose RFKG-CoT: First, it replaces the rigid hop-count selector with a relation-driven adaptive hop-count selector that dynamically adjusts reasoning steps by activating KG relations (e.g., 1-hop for direct "brother" relations, 2-hop for indirect "father-son" chains), formalized via a relation mask. Second, it introduces a few-shot in-context learning path guidance mechanism with CoT (think) that constructs examples in a "question-paths-answer" format to enhance LLMs' ability to understand reasoning paths. Experiments on four KGQA benchmarks show RFKG-CoT improves accuracy by up to 14.7 pp (Llama2-7B on WebQSP) over KG-CoT. Ablations confirm the hop-count selector and the path prompt are complementary, jointly transforming KG evidence into more faithful answers.

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。

↑