摘要
目前网络上关于辽代历史信息化智能查询服务系统资源相对缺乏,关于辽代历史介绍文本篇幅冗长,不方便大众进行阅读观看。为了便于大众能更加快速准确了解相关的辽代历史知识,基于百度百科、搜狗百科以及基于爬虫技术等形式获取了与辽代历史相关的文本数据,采用BiLSTM-CRF模型进行实体抽取,通过关系抽取模型抽取实体间的关系,通过知识融合等技术对得到的数据进行实体对齐以及本体构建。最终构建辽代历史文化领域知识图谱,并在此知识图谱的基础上开发了可视化查询系统。
It is found that the resources of the intelligent query service system about the historical information of the Liao Dynasty on the Internet are relatively lacking.At the same time,the texts on the Internet about the history of the Liao Dynasty are relatively lengthy and inconvenient for the public to read.In order to facilitate the public to understand the relevant historical knowledge of the Liao Dynasty more quickly and accurately,this study first obtains text data related to the history of the Liao Dynasty based on Baidu Encyclopedia,Sogou Encyclopedia,and crawler technology.Next,the BiLSTM-CRF model is used for entity extraction,and the relationship between entities is extracted through the relationship extraction model.Then,entity alignment and ontology construction are performed on the obtained data through knowledge fusion technology.Finally,a knowledge graph of the history and culture of the Liao Dynasty is constructed,and a visual query system is developed on the basis of this knowledge graph.
作者
刘爽
谭楠楠
杨辉
LIU Shuang;TAN Nan-nan;YANG Hui(School of Computer Science and Engineering, Dalian Minzu University, Dalian Liaoning 116650, China)
出处
《大连民族大学学报》
2021年第1期73-80,共8页
Journal of Dalian Minzu University
基金
辽宁省经济社会发展研究课题(2021lslybkt-022)。
关键词
辽代历史文化
知识图谱
知识抽取
命名实体识别
history and culture of the Liao Dynasty
knowledge graph
knowledge extraction
named entity recognition