摘要
知识图谱因其所表示知识的精准性和特有的图结构而广泛应用于工业领域。工业知识图谱通常在领域本体库中以领域数据为导向获取本体子集,从而生成对应的数据模式。针对工业知识图谱搭建的问题,基于本体重构的方法自动化了将给定领域本体转换为数据模式的整个过程。首先,在已有本体重构算法中加入多模块匹配机制,以降低重构过程中的人力成本;其次,规范了在工业领域搭建知识图谱的整体逻辑框架;最后,使用工业领域真实数据,构建了认证机构领域的知识图谱框架,并与未使用该框架的其他构建方式进行对比,证明了所提出方法的可行性。
Knowledge graph is widely used in industrial due to its precision in representing knowledge and unique graph structure.Industrial knowledge graph usually obtains ontology subsets in domain ontology libraries guided by domain data,thus generates corresponding data patterns.For the construction of industrial knowledge graph,the ontology reconstruction method automates the entire process of transforming a given domain ontology into a data pattern.First,the multi-module matching mechanism is added to the existing ontology reconstruction algorithm to reduce the labor cost in the reconstruction process.Then,the overall logical framework for constructing a knowledge graph in industrial is standardized.Finally,the knowledge graph framework in certification authority domain is constructed based on real data in industrial,and compared with other construction methods that do not use the framework,the feasibility of the proposed method is verified.
作者
翟值楚
靖琦东
滕磊
李倩
ZHAI Zhichu;JING Qidong;TENG Lei;LI Qian(CEC Industrial Internet Co.,Ltd.,Changsha Hunan 410205,China)
出处
《通信技术》
2023年第5期620-625,共6页
Communications Technology
关键词
工业知识图谱
本体重构
关系映射
工业数据分析
industrial knowledge graph
ontology reconstruction
relationship mapping
industrial data analysis
