BACKGROUND
To combat the COVID-19 pandemic, various vaccines have been developed and their rollout is under way. However, the uptake rate is hindered by vaccine hesitancy influenced by the conversations on social media. It is necessary to trace public opinion toward COVID-vaccines on social media.
OBJECTIVE
The objective of this study is to examine the sentiments and topics of English-language twitter discussion regarding COVID-19 vaccination. Further this study also aims to explore the temporal trend of sentiments and topics over one month in the early period of vaccine roll-out.
METHODS
Following existing studies of vaccine acceptance and social media, we collected Tweet posts from Twitter data base using Twitter API from December 2020 to January 2021, which reflected actual public discussions toward COVID vaccination after the beginning of the rollout. After data cleansing and selection, 656,102 vaccine-related tweets were identified from 329,441 unique users. We leveraged VADER (Valence Aware Dictionary and sEntiment Reasoner) sentiment analysis tool to explore sentiment scores and Natural Language Toolkit (NLTK) to confirm relevant topics. We also depicted daily changes of sentiments and topics in COVID-vaccine-related tweets across one month period.
RESULTS
Forty-two percentage of tweets expressed pro-vaccine sentiments while 21% held negative attitudes. The trend of sentiment kept positive and consistent overtime, but a sudden surge of negative tweets occurred around the New Year, which was caused by some unexpected adverse incidents. The Six main topics associated with vaccines were identified: Advocation of vaccination (42,459, 6.47%), Official information releases (29,847, 4.55%), Vaccine distribution (12,946, 1.97%), Vaccine safety concerns (11,236, 1.71%), Personal vaccination experience (5,594, 0.85%) and Conspiracy theory (2,962, 0.45%). Among popular tweets that have been reposted frequently, adverse incidents reported by reliable source have triggered intense discussions about vaccine safety issues, usually in a negative attitude.
CONCLUSIONS
: Most tweets expressed non-negative sentiments toward vaccination. However, vaccination-related adverse incidents have triggered intense discussions in a negative attitude. Our findings can help policymakers and health providers view the whole picture of the influence of social media and develop better communicative strategies for improving vaccine acceptance.