用大模型融合多语言知识图谱，解决跨语言语义差异问题
Graph Fusion Across Languages using Large Language Models
- 将三元组转为自然语言序列，让大模型理解并对齐跨语言图结构
- 在DBP15K上实现多图连续融合，成功处理异构数据
- 适合需要多源多语言知识整合的研究者与系统开发者
跨语言知识图谱的融合因语义异质性和图结构复杂性而长期面临挑战。本文提出一种基于大语言模型（LLM）的跨语言图谱融合框架，利用其上下文推理能力与多语言语义先验。通过将三元组直接映射为自然语言序列（如[头][关系][尾]），使LLM能够对齐关系和实体，实现对演进中的融合图 $G_{c}^{(t-1)}$ 与新候选图 $G_{t}$ 的动态整合。在DBP15K数据集上的实验表明，大模型可作为通用语义桥梁，有效解决跨语言歧义问题。结果展示了多个异构图的有序聚合，提供了一种可扩展、模块化的多源多语言持续知识合成方案。
原文摘要 · Abstract (English)
Combining multiple knowledge graphs (KGs) across linguistic boundaries is a persistent challenge due to semantic heterogeneity and the complexity of graph environments. We propose a framework for cross-lingual graph fusion, leveraging the in-context reasoning and multilingual semantic priors of Large Language Models (LLMs). The framework implements structural linearization by mapping triplets directly into natural language sequences (e.g., [head] [relation] [tail]), enabling the LLM to map relations and reconcile entities between an evolving fused graph ($G_{c}^{(t-1)}$) and a new candidate graph ($G_{t}$). Evaluated on the DBP15K dataset, this exploratory study demonstrates that LLMs can serve as a universal semantic bridge to resolve cross-lingual discrepancies. Results show the successful sequential agglomeration of multiple heterogeneous graphs, offering a scalable, modular solution for continuous knowledge synthesis in multi-source, multilingual environments.
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。