Disaster Impacts Surveillance from Social Media with Topic Modeling and Feature Extraction: Case of Hurricane Harvey

作     者：Volodymyr V.Mihunov Navid H.Jafari Kejin Wang Nina S.N.Lam Dylan Govender Volodymyr V.Mihunov;Navid H.Jafari;Kejin Wang;Nina S.N.Lam;Dylan Govender

作者机构：Department of Environmental SciencesLouisiana State UniversityBaton RougeLA 70803USA Department of Civil and Environmental EngineeringLouisiana State UniversityBaton RougeLA 70803USA Division of Electrical and Computer EngineeringLouisiana State UniversityBaton RougeLA 70803USA 

出 版 物：《International Journal of Disaster Risk Science》 (国际灾害风险科学学报（英文版）)

年 卷 期：2022年第13卷第5期

页      面：729-742页

核心收录：

学科分类：083002[工学-环境工程] 0830[工学-环境科学与工程（可授工学、理学、农学学位）] 07[理学] 070601[理学-气象学] 08[工学] 0706[理学-大气科学] 

基　　金：This article is based on work supported by two grants from the National Science Foundation of the United States(under Grant Numbers 1620451 and 1945787).Any opinions fndings and conclusions or recommendations expressed in this article are those of the authors and do not necessarily refect the views of the National Science Foundation 

主　　题：Disaster impacts Hurricane Harvey Infrastructure impacts Latent Dirichlet allocation(LDA) Social media analysis Twitter data 

摘      要：Twitter can supply useful information on infrastructure impacts to the emergency managers during major disasters,but it is time consuming to filter through many irrelevant *** studies have identified the types of messages that can be found on social media during disasters,but few solutions have been proposed to efficiently extract useful *** present a framework that can be applied in a timely manner to provide disaster impact information sourced from social *** framework is tested on a well-studied and data-rich case of Hurricane *** procedures consist of filtering the raw Twitter data based on keywords,location,and tweet attributes,and then applying the latent Dirichlet allocation(LDA) to separate the tweets from the disaster affected area into categories(topics) useful to emergency *** LDA revealed that out of 24 topics found in the data,nine were directly related to disaster impacts-for example,outages,closures,flooded roads,and damaged *** such as frequent hashtags,mentions,URLs,and useful images were then extracted and *** relevant tweets,along with useful images,were correlated at the county level with flood depth,distributed disaster aid(damage),and population *** correlations were found between the nine relevant topics and population density but not flood depth and damage,suggesting that more research into the suitability of social media data for disaster impacts modeling is *** results from this study provide baseline information for such efforts in the future.