动态调整推理步数并引导路径，提升大模型问答准确性
RFKG-CoT: Relation-Driven Adaptive Hop-count Selection and Few-Shot Path Guidance for Knowledge-Aware QA
- 根据关系类型自适应选择推理步数，避免固定跳数限制
- 在WebQSP上相比KG-CoT最高提升14.7个百分点准确率
- 适合需要高可靠知识问答的场景，如医疗或金融领域
大语言模型在知识密集型问答中常因参数化知识局限产生幻觉。现有方法如KG-CoT通过整合知识图谱路径提升可靠性，但存在跳数选择僵化（仅由问题驱动）和推理路径利用不足（缺乏引导）的问题。为此，我们提出RFKG-CoT：首先，用关系驱动的自适应跳数选择器替代固定跳数机制，通过关系掩码动态调整推理步骤（例如“兄弟”关系用1跳，“父子”链用2跳）；其次，引入少样本上下文学习路径引导机制，以“问题-路径-答案”格式构造思维链示例，增强模型对推理路径的理解。在四个KGQA基准上的实验表明，RFKG-CoT在Llama2-7B模型上对WebQSP的准确率最高提升14.7个百分点。消融实验验证了跳数选择器与路径提示具有互补性，共同将知识图谱证据转化为更可信的答案。
原文摘要 · Abstract (English)
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.
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。