用检索增强与词汇约束提升大模型构建知识图谱的准确性
pro-team at LLMs4OL 2026 Tasks Flagship and Reuse: Retrieval-Augmented Generation and Vocabulary-Constrained Filtering for Ontology Learning
- 通过检索相似案例并限制生成词汇，减少大模型幻觉
- 在任务B上达到0.92的术语分类准确率，图谱相似度达0.8692
- 适合需要高精度术语和层级关系提取的应用场景
尽管大语言模型在文本到知识图谱的构建中取得进展，仍存在术语幻觉、格式不一致及偏好层级关系等问题。在LLMs4OL 2026挑战赛中，我们针对端到端旗舰任务（任务A）和知识图谱扩展复用任务（任务B），采用离线检索增强少样本提示方法。系统使用Qwen2.5-14B-Instruct搭配all-MiniLM-L6-v2进行示例检索，任务A选取前5个示例，任务B取前2个。采用左截断上下文窗口策略，确保任务指令在长提示中保留。任务B生成三元组后执行确定性词汇约束过滤：仅保留至少一端属于样本封闭词汇表的三元组，并移除初始图谱的重复项。该方法在任务B上获得0.8692的语义图相似度、0.9200的术语分类F1值和0.8540的本体发现F1值；任务A得分为0.7416。但未提取非层级关系，说明封闭的层级导向关系词表存在局限。
原文摘要 · Abstract (English)
Ontology learning from text remains challenging despite significant progress in Large Language Models (LLMs), which can hallucinate domain terms, produce inconsistent formats, and favor hierarchical over associative relations. In the LLMs4OL 2026 Challenge, we address both the End-to-End Flagship Task (Task A) and Ontology Extension Reuse Task (Task B) using an offline retrieval-augmented few-shot prompting pipeline. Our system employs Qwen2.5-14B-Instruct with all-MiniLM-L6-v2 for demonstration retrieval, selecting the top-5 examples for Task A and top-2 for Task B. A left-truncated context-windowing strategy preserves task instructions within long prompts. For Task B, generated triples undergo deterministic vocabulary-constrained filtering, retaining triples when at least one endpoint belongs to the sample's closed term/type vocabulary and removing duplicates of the initial ontology. The approach achieves Semantic Graph Similarity of 0.8692, Term-Typing F1 of 0.9200, and Taxonomy Discovery F1 of 0.8540 on Task B, while Task A achieves 0.7416 Semantic Graph Similarity. However, no non-taxonomic relations are extracted, highlighting limitations of closed, taxonomy-oriented relation vocabularies.
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。