Emotional communication analysis of emergency microblog based on the evolution life cycle of public opinion

Authors:
Publisher website
Google Scholar
Abstract
Purpose: Microblogging is an important channel used to disseminate online public opinion during an emergency. Analyzing the features and evolution mechanism of online public opinion during an emergency plays a significant role in crisis management. Design/methodology/approach: This paper uses the event of Hurricane Irma and combines it with the life cycle of online public opinion evolution to understand the effect of different types of emotional (joy, anger, sadness, fear, disgust) microblogs (tweets) on information dissemination. The research was performed in the context of Hurricane Irma by using tweets associated with that event. Findings: This paper demonstrates that negative emotional information has a greater communication effect, and further, the target audience that receives more exposure to negative emotional microblogs has a stronger tendency to retweet. Meanwhile, emotions expressed in tweets and the life cycle of public opinion evolution exert interactive effects on the retweeting behavior of the target audience. Research limitations/implications: For future research, a professional dictionary and the context should be taken into consideration to make the modeling in the text more normative and analyzable. Practical implications: This paper aims to reveal how the emotions of a tweet affect its virality in terms of diffusion volume in the context of an emergency event. Social implications: The conclusion made in this paper can shed light on the real-time regulation and public opinion transmission, as well as for efficient intelligence service and emergency management. Originality/value: In this study, Hurricane Irma is taken as an example to explore the factors influencing the information dissemination during emergencies on the social media environment. The relationship between the sentiment of a tweet and the life cycle of public opinion and its effect on tweet volume were investigated.

This publication has 17 references indexed in Scilit: