COVID-19 vaccines in Twitter ecosystem: Analyzing perceptions and attitudes by sentiment and text analysis method


Kahraman E., Demirel S., Gündüz U.

Journal of Public Health: From Theory to Practice, cilt.1, sa.1, ss.1-15, 2023 (SSCI)

Özet

Aim An abundance of information and rumors pertaining to COVID-19 vaccines has disseminated extensively especially through Twitter. The primary objective of this study is to explore and analyze the prevailing perceptions and attitudes towards COVID-19 vaccines as manifested within the Twitter ecosystem.

Subject and Methods The utilization of social media platforms for conducting public health analyses during pandemics has garnered heightened attention. Twitter, in particular, offers the potential to present trustworthy and real-time data regarding public opinions during crises, owing to the presence of verified accounts belonging to public health officials and authorities. This study employs a text mining methodology and sentiment analysis to examine a comprehensive dataset comprising 66,048 tweets. These tweets, posted between the 5th and 14th of October 2021, focus on four COVID-19 vaccines (AstraZeneca, Biontech, Sinovac and Sputnik5), with the aim of scrutinizing the prevailing perceptions and attitudes towards these vaccines within the Twitter community.

Results The results are presented as text and sentiment analysis. As a result of the text analysis, the efficacy and side effects of the vaccines are the main topics to be discussed. According to sentiment analysis, AstraZeneca and Biontech have more percentage of negative tweets associated with them whereas Sinovac and Sputnik5 have more percentage of positive tweets.

Conclusion The sentiment analysis of tweets regarding vaccines highlights the intricate relationship between the textual aspects and formal features of the tweets. Furthermore, it offers insights into the level of influence and dissemination exhibited by these tweets within the Twitter ecosystem.