We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Sentiment Analysis of Rumor Spread Amid COVID-19: Based on Weibo Text.
- Authors
Wang, Peng; Shi, Huimin; Wu, Xiaojie; Jiao, Longzhen
- Abstract
(1) Background: in early 2020, COVID-19 broke out. Driven by people's psychology of conformity, panic, group polarization, etc., various rumors appeared and spread wildly, and the Internet became a hotbed of rumors. (2) Methods: the study selected Weibo as the research media, using topic models, time series analysis, sentiment analysis, and Granger causality testing methods to analyze the social media texts related to COVID-19 rumors. (3) Results: in study 1, we obtained 21 topics related to "COVID-19 rumors" and "outbreak rumors" after conducting topic model analysis on Weibo texts; in study 2, we explored the emotional changes of netizens before and after rumor dispelling information was released and found people's positive emotions first declined and then rose; in study 3, we also explored the emotional changes of netizens before and after the "Wuhan lockdown" event and found positive sentiment of people in non-Wuhan areas increased, while negative sentiment of people in Wuhan increased; in study 4, we studied the relationship between rumor spread and emotional polarity and found negative sentiment and rumor spread was causally interrelated. (4) Conclusion: These findings could help us to intuitively understand the impact of rumors spread on people's emotions during the COVID-19 pandemic and help the government take measures to reduce panic.
- Subjects
WUHAN (China); GRANGER causality test; COVID-19; SENTIMENT analysis; COVID-19 pandemic; RUMOR; TIME series analysis; VECTOR error-correction models
- Publication
Healthcare (2227-9032), 2021, Vol 9, Issue 10, p1275
- ISSN
2227-9032
- Publication type
Article
- DOI
10.3390/healthcare9101275