基于政务微博的自然灾害知识图谱构建——以森林火灾为例

Construction of Natural Disaster Knowledge Map Based on Government Microblog——Taking Forest Fires as an Example

[目的/意义]利用政务微博信息构建自然灾害知识图谱,旨在为相关部门加强自然灾害事件的管理提供知识层面的参考价值。[方法/过程]以森林火灾事件为例,选取政务微博信息资源,首先使用LDA主题模型划分微博资源主题;其次构建自然灾害知识图谱的模式层和数据层,包括本体构建、实体抽取、关系抽取和数据融合等环节;最后使用Neo4j图数据库实现自然灾害知识图谱的存储与检索,实现自然灾害信息的组织与可视化。[结果/结论]构建了基于主题划分的自然灾害知识图谱,实现了自然灾害信息的主题知识关联与规范化处理,对提升政府部门防范与管理自然灾害的科学决策水平具有积极作用。

[Purpose/Significance]This paper constructs natural disaster knowledge mapping using information from governmental microblogs,aiming to provide a knowledge level reference value for relevant departments to strengthen the management of natural disaster events.[Method/Process]Taking the forest fire incident as an example,this paper selected microblog information resources and firstly used the LDA topic model to divide the microblog resource topics;secondly,constructed the natural disaster knowledge graph schema layer and data layer,including ontology construction,entity extraction,relationship extraction and data fusion;finally,used the Neo4j graph database to realize the presentation of the natural disaster knowledge graph,and organizes and visualization.[Results/Conclusions]This paper constructs a natural disaster knowledge graph based on topic division,realizes the topic knowledge association and normalization of natural disaster information,and improve the scientific decision-making level of government departments to prevent and manage natural disasters.