摘要
知识图谱技术在科学统筹装备数据、整合繁杂的装备信息上有较好的效果。为解决通用知识图谱对装备实体的覆盖率不高、装备命名实体不规范的问题,提出了一种军事装备概念图谱构建的方法。该方法依托现有兵器库的结构化数据构建装备概念图谱,采用迭代学习的方法,以开放的多源数据为基础进行了装备实体的补全,确保了概念图谱的广度和精度。该技术不仅有助于提高机器对于装备的认知能力,也可作为本体库应用于装备知识图谱的构建。
Knowledge graph has a good effect on scientific coordination of equipment data and integration of complex equipment information.In view of the problems of the low coverage rate of general-purpose knowledge graph to equipment entities and the irregularity of equipment named entity,a method for constructing equipment concept graph is proposed.Building on the existing arsenal of structured data,this approach is using an iterative learning method based on open multi-source data for equipment entity completion to ensure the breadth and precision of the concept graph.This technology not only helps to improve the cognitive ability of machines for equipment but also can be used as an ontology library in the construction of equipment knowledge graph.
作者
姚奕
杨帆
刘语婵
袁清波
YAO Yi;YANG Fan;LIU Yu-chan;YUAN Qing-bo(Command and Control Engineering College,Army Engineering University,Nanjing 210007,China)
出处
《火力与指挥控制》
CSCD
北大核心
2021年第9期125-132,共8页
Fire Control & Command Control
关键词
知识图谱
概念图谱
军事装备
ISA
关系抽取
关系补全
konwledge graph
concept graph
military equipment
isA relationship extraction
relation completion