情报科学 ›› 2021, Vol. 39 ›› Issue (1): 72-80.

• 理论研究 • 上一篇    下一篇

基于网络新闻语料的公共危机事件知识图谱构建 

  

  • 出版日期:2021-01-01 发布日期:2021-01-25

  • Online:2021-01-01 Published:2021-01-25

摘要: 【目的/意义】本文提出了借助知识图谱的知识互联和可视化分析等功能,为公共危机事件智能治理提供知
识支撑的观点。【方法
/过程】以“港独”为例,从新华网、人民网、中国新闻网等“港独”新闻语料中抽取实体,构建公
共危机事件知识图谱模型。从知识图谱本体构建和实体数据填充两个方面,对“港独”知识图谱的实体抽取、关系
抽取、知识融合、知识加工和知识推理等构建环节进行重点论述和实践,最后,利用
Neo4j数据库实现知识存储,
成的“港独”知识图谱。【结果
/结论】通过Cypher语句对“港独”知识图谱进行查询,能够呈现可视化语义网结构图,
利于知识隐重发现。【创新
/局限】本文的创新之处在于把知识图谱技术应用于公共危机事件治理中,为其智能查
询和决策分析提供数据支持。在今后实践中,数据源的选取不必局限于网络新闻语料,还可以适当对“港独”人员
的微博、微信、论坛中的信息进行采集和抽取,以此扩大知识关联,从而更为全面、高效地进行公共危机事件智能
治理。

Abstract: 【Purpose/significance】This paper puts forward the view of knowledge support for intelligent governance of public crisis
events by means of Knowledge Mapping.【Method/process】Taking "Hong Kong Independence" as an example, the entities are extract⁃
ed from Xinhua Net,People’s Daily Online and China News Net, and knowledge mapping model is drawn. The entity extraction, rela⁃
tionships extraction, knowledge fusion etc are conducted. Finally, the knowledge map of "Hong Kong Independence" is generated by
knowledge storage using Neo4j database.【Result/conclusion】By querying the knowledge map with Cypher, the semantic web struc⁃
ture of "Hong Kong Independence" is visualized,more hidden knowledge is discoveried.【Innovation/limitation】The innovation of this
paper lies in the application of knowledge mapping technology in public crisis event management ,it provides data support for intelli⁃
gent query and decision analysis. In future practice, the selection of data sources need not be limited to the network news corpus, the
information from the micro-blog, WeChat and forum of the "Hong Kong Independence" personnel can be collected for expanding the
knowledge connection and improving intelligent management of the public crisis event.