Understanding public opinion in different disaster stages: a case study of Hurricane Irma

Abstract

Purpose Social media, such as Twitter, has become the first and the most frequent place to visit in order to gain information and establish situational awareness in emergencies and disasters. The purpose of this paper is to examine public opinion on Twitter in different disaster stages using the case of Hurricane Irma. Design/methodology/approach More than 3.5m tweets capturing the entire disaster lifecycle were collected and analyzed. Topic modeling was used to generate topics at each disaster stage based on Fink’s (1986) four-stage model of crisis and disaster: prodromal, acute, chronic and termination stages. Findings The results revealed that media reliance varied across different stages. All topics in the prodromal stage were associated with the early warning and real-time news. The topic of lessons learned from Hurricane Harvey was the most popular at this stage. The acute stage recorded the highest number of daily tweets. The most popular topic was the safety of people and animals. In the chronic stage too, the safety of people and animals remained a major concern. Heroic and anti-social behaviors also received substantial attention. In the termination stage, climate change was the most frequently discussed topic. Politics-related discussions were heated. Originality/value The results extended and enhanced the four-stage model of crisis and disaster. These findings can help government agencies and crisis managers address audience needs effectively at various crisis stages in a timely manner.

Temporal, Spatial, and Socioeconomic Dynamics in Social Media Thematic Emphases during Typhoon Mangkhut

“…In the early hours of the disaster, Theme 3 (typhoon status and impact) is the most popular theme, but its popularity soon fell sharply after. According to Fink's (1986) four-stage model of disaster [48], Xu et al [4] found that themes varied across different stages during Hurricane Irma. Zhao et al [5] also confirmed that the diversity of major themes was significantly associated with the crisis stages.…”

“…Zhao et al [6] attempted to understand social media publics' changing concerns by analyzing whether and how the publics share messages of different themes and forms in different stages of a crisis. Based on a four-stage model of disaster, Xu et al [4] found different themes on Twitter in different disaster stages using the case of Hurricane Irma.…”

“…Narratives regarding a disaster often consist of several themes or topics, and each theme or topic allows people to understand and communicate from a certain perspective [4]. Theme evolution can be detected in different stages of a disaster in order to understand the users' concerns of social media at different stages of a crisis [4][5][6].…”

Identifying and Analyzing Health-Related Themes in Disinformation Shared by Conservative and Liberal Russian Trolls on Twitter

“…Among topic models, Latent Dirichlet Allocation (LDA) [ 31 ] is an effective model [ 32 ] that has been utilized for a wide range of domains, such as understanding public opinion [ 29 , 33 ], analysis of health documents [ 34 , 35 ], mining online reviews [ 36 , 37 ], and developing a systematic literature review [ 38 , 39 , 40 ]. LDA has also been used to characterize social media discussions on different issues, such as diet [ 41 ], exercise [ 42 ], LGBT health [ 43 , 44 ], antiquarantine discussions during the COVID-19 pandemic [ 45 ], politics [ 46 ], and natural disasters [ 47 , 48 ]. LDA creates topics including the probability of each word (W) given a topic (T) or P(W|T).…”

News Frames and News Exposure Predicting Flu Vaccination Uptake: Evidence from U.S. Newspapers, 2011–2018 Using Computational Methods

“…The present study focuses on topical frames, which are determined by topic modeling, a machine learning technique to assign unstructured information to different topics fully automatically (Xu et al, 2019). As one of the most frequently applied and validated methods (Deveaud et al, 2014), Latent Dirichlet Allocation (LDA) topic modeling is based on the Bayesian hierarchical model.…”