基于多模态大模型的知识图谱进化研究  

Research on knowledge graph evolution based on multimodal large models

作　　者：王思博 WANG Sibo(CRRC Information Technology Co.,Ltd.,Beijing 100039,China)

机构地区：[1]中车信息技术有限公司,北京100039

出　　处：《中国高新科技》2025年第2期25-27,共3页

摘　　要：针对传统知识图谱存在的信息不完整、数据稀疏、更新困难等问题,提出了一种基于多模态大模型的知识图谱进化方法,采用了一种迁移学习的思想,将已有的知识图谱作为先验知识,利用多模态大模型从未标注的数据中学习新的实体和关系。实践证明,基于多模态大模型的知识图谱进化方法能显著提高知识图谱的完整性,与传统方法相比,该方法在信息抽取和知识推理任务上取得了更好的效果。Traditional knowledge maps have some problems,such as incomplete information,sparse data,and difficulty in updating.In order to solve these problems,an evolutionary method of knowledge maps based on multimodal large model is proposed,which adopts the idea of transfer learning,takes existing knowledge maps as prior knowledge,and uses multimodal large model to learn new entities and relationships from unlabeled data.Practice has proved that the knowledge map evolution method based on multi-modal large model can significantly improve the integrity of knowledge map.Compared with traditional methods,this method has achieved better results in information extraction and knowledge reasoning tasks.

关 键 词：多模态大模型 知识图谱 进化 应用 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]