摘要
以2020年第6号台风“米克拉”为例,采集了5916条新浪微博作为数据源,综合应用隐含狄利克雷分布(Latent Dirichlet Allocation,LDA)主题模型、文本情感分析方法和空间分析技术,挖掘、分析台风的灾情时空过程。基于LDA主题模型建立了主题-词矩阵并进行隐含主题聚类,这些微博文本被分为灾损类信息、预警类信息、防御类信息和无关信息;从主题信息和文本情感值两个角度入手,对此次台风事件网络舆情的演化过程进行分析。结果表明:“米克拉”登陆前有大量积极情感的微博,主要包含渴望降雨、降温等信息,此类微博大量分布在漳州、厦门、福州等地区;在台风入境后消极情感的微博大量增多,主要描述道路、树木等受大风和大雨影响的灾损类信息,此类信息的空间位置主要分布在漳州和厦门,能较好地反映台风灾害影响的时空分布。通过对微博主题类别和情感极性进行时空分析,实现台风灾害事件发展趋势的监测,为防灾减灾提供参考依据。
While the No.6 typhoon“Mekkhala”appeared and landed at Zhangpu,Fujian Province,P.R.China,from Aug.8th to 13th,2020,5916 records of Sina Weibo about the tropical cyclone were collected as a dataset to explore the disaster process.To analyze its temporal and spatial process,Latent Dirichlet Allocation(LDA)topic model,text sentiment analysis method and spatial analysis technology were employed for data mining.Based on a matrix of keywords and themes implicit clustering of LDA model,all Weibo texts were divided into four topics as flood-loss,early warning,defense and irrelevant information;and then the influence of the typhoon events on the Internet public opinion was measured with two sides,keywords and emotional values.The results showed that,there were a lot of microblogging with positive emotions before“Mekkhala”landing,since people desire for a raining and cooling weather after the long heat days.After the typhoon landed,a large number of negative emotion microblogs were increased which mainly described all kinds of damage on roads,trees and other disaster-related information affected by strong wind and heavy rain.The location of microblogs was mainly around Zhangzhou and Xiamen,which reflects the impact of temporal and spatial distribution of typhoon disasters very well.The development trend of typhoon disaster events can be monitored through the temporal and spatial analysis of Weibo subject categories and emotional polarity,which could provide reference for disaster prevention and mitigation.
作者
陈齐超
林广发
梁春阳
黄潇
张明锋
陈鑫
周星辰
CHEN Qichao;LIN Guangfa;LIANG Chunyang;HUANG Xiao;ZHANG Mingfeng;CHEN Xin;ZHOU Xingchen(School of Geographical Sciences, Fujian Normal University, Fuzhou 350007, China;Fujian Provincial Engineering Research Center for Monitoring and Assessing Terrestrial Disasters, Fuzhou 350007, China;Research Center for National Geographical Condition Monitoring and Emergency Support in the Economic Zone on the West Side of the Taiwan Strait, Fuzhou 350007, China)
出处
《亚热带资源与环境学报》
2021年第1期70-76,共7页
Journal of Subtropical Resources and Environment
基金
国家重点研发计划重点专项(2016YFC0502905)
福建省公益类科研院所专项(2015R1034-1)。
关键词
新浪微博
台风灾害
主题模型
情感分析
灾情评估
Sina Weibo
typhoon disaster
topic model
sentiment analysis
disaster assessment
