基于震后舆情的灾情信息提取研究  

Research on Disaster Information ExtractionBased on Post-Earthquake Public Sentiment

作　　者：闫晓美[1,2] 牛艳杰 王宁 许振鹏 YAN Xiaomei;NIU Yanjie;WANG Ning;XU Zhenpeng(Shanxi Earthquake Agency,Taiyuan,Shanxi 030021,China;National Continental Rift Valley Dynamics Observatory of Taiyuan,Taiyuan,Shanxi 030025,China)

机构地区：[1]山西省地震局,山西太原030021 [2]太原大陆裂谷动力学国家野外科学观测研究站,山西太原030025

出　　处：《山西地震》2024年第1期1-4,16,共5页Earthquake Research in Shanxi

基　　金：山西省地震局青年项目(SBK-2312)。

摘　　要：大地震发生后,通常会出现大量关于地震灾害的信息并在网络中快速传播,为快速准确地获取地震灾情信息,开展基于震后舆情提取灾情信息的相关研究。首先构建灾情信息挖掘模型,建立灾情关键词表,通过计算词向量相似度,快速提取地震灾情等相关信息。同时以“山东德州原平5.5级地震”为案例,验证构建模型在地震灾情提取方面的效果及其可行性和实用性。结果表明,该研究可为震后快速应对和处置灾情提供一定的参考数据和决策依据。After a major earthquake,a large amount of information regarding earthquake disasters usually emerges and spreads rapidly on the internet.To obtain earthquake disaster information quickly and accurately,this study carries out research on extracting disaster information based on post-earthquake public sentiment.First,a disaster information mining model is constructed,and a disaster keyword list is established.By calculating word vector similarity,earthquake disaster information and related information are quickly extracted.At the same time,the"5.5 magnitude earthquake in Yuanping,Dezhou,Shandong"case is used to verify the effectiveness,feasibility,and practicality of the constructed model in extracting earthquake disaster information.The results show that this research can provide certain reference data and a decision-making basis for rapid response and disposal of disaster situations after an earthquake.

关 键 词：地震灾情 震后舆情 网络爬虫 文本挖掘 信息提取 

分 类 号：P315-39[天文地球—地震学]