基于CiteSpace的国内命名实体识别技术的知识图谱分析  

Knowledge Graph Analysis of Domestic Named Entity Recognition Based on CiteSpace

作　　者：李源 蔡忠祥 李娜 黄子鸣 LI Yuan;CAI Zhongxiang;LI Na;HUANG Ziming(Faculty of Information Engineering,Xinyang Agriculture and Forestry University,Xinyang 464000,China)

机构地区：[1]信阳农林学院信息工程学院,河南信阳464000

出　　处：《现代信息科技》2024年第15期124-128,133,共6页Modern Information Technology

基　　金：信阳农林学院青年教师科研基金(QN2023014)。

摘　　要：命名实体识别(NER)是自然语言处理(NLP)领域的重要任务。量化梳理NER的文献进程,有助于未来NER乃至NLP技术的突破发展。目前已有大量学者对NER任务进行了综述回顾。但是这些基于传统文献计量学的回顾方式,既过于依赖专家经验,又无法直观呈现知识变迁。因此,文章以知网中文核心文献为驱动,基于CiteSpace工具,对国内NER技术进行了知识图谱展示和数据挖掘分析。分析结果显示:以2016年为界,国内已形成基于数据驱动的两个快速发展期;“当地高校+省立实验室”的形式已成为国内机构合作的主流;“深度学习”“知识图谱”“实体识别”“信息抽取”“神经网络”和“词向量”是国内NER领域的研究热点;未来对NER的研究倾向于应用落地、数据增强和知识抽取。Named Entity Recognition(NER)is an important task in the field of Natural Language Processing(NLP).Quantitative review of the literature process of NER technology is conducive to the breakthrough development of NER and even NLP technology in the future.At present,a large number of scholars have reviewed the NER tasks.However,these review methods based on traditional bibliometrics not only rely too much on expert experience,but also can not directly present the change of knowledge paradigm.Therefore,this paper,driven by the Chinese core literature of CNKI and based on CiteSpace,presents the Knowledge Graph display and data mining analysis of domestic NER technology.The analysis results show that the domestic research on NER has formed two rapid development periods based on data drive from 2016.The form of“local university+provincial laboratory”has become the mainstream of domestic institutional cooperation.“Deep Learning”“Knowledge Graph”“entity recognition”“information extraction”“Neural Network”and“word vector”are the hot spots of domestic NER research.Future research on NER tends to apply landing,data enhancement and knowledge extraction.

关 键 词：命名实体识别 CITESPACE 知识图谱 中国知网 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]