基于知识图谱的舰船信息分析识别技术研究  

Research on Ship Information Analysis and Identification Technology Based on Knowledge Graph

作　　者：薛亮 周游 XUE Liang;ZHOU You(No.91977 Troops of PLA,Beijing 100000)

机构地区：[1]中国人民解放军91977部队,北京100000

出　　处：《舰船电子工程》2021年第10期48-51,145,共5页Ship Electronic Engineering

摘　　要：论文提出了一个基于知识图谱的舰船信息分析识别模型和算法,该模型采用基于BERT+Bi-LSTM+CRF的命名实体识别技术和基于Siamese+BERT+Bi-LSTM的属性映射技术,基于该模型和算法,实现了基于知识图谱的舰船信息分析识别的原理系统,并选用舰船领域知识图谱作为知识库进行了仿真实验,实验表明,论文提出的系统可以达到较高的准确率。A model and algorithm of ship information analysis and recognition based on knowledge map is proposed in this paper.The model uses named entity recognition technology based on BERT+Bi-LSTM+CRF and attribute mapping technology based on Siamese+BERT+Bi-LSTM.Based on the model and algorithm,the principle system of ship information analysis and recognition based on knowledge map is realized.A simulation experiment is carried out by choosing the knowledge map of the ship domain as the knowledge base.The simulation results show that the proposed system can achieve high accuracy.

关 键 词：信息分析识别 知识图谱 实体识别 属性映射 

分 类 号：R222.18[医药卫生—中医基础理论]