How was the coronavirus vaccine accepted on Twitter? A computational analysis using big data in Japan

Abstract

Purpose

This study aims to reveal how the COVID-19 vaccine was accepted in the Japanese Twitter-sphere. This study explores how the topics related to the vaccine promotion project changed on Twitter and how the topics that were likely to spread changed during the vaccine promotion project.

Design/methodology/approach

The computational social science methodology was adopted. This study collected all tweets containing the word “vaccine” using the Twitter API from March to October 2021 and conducted the following analysis: analyzing frequent words and identifying topics likely to spread through the cosine similarity and Tobit model.

Findings

First, vaccine hesitancy–related words were frequently mentioned during the vaccine introduction and dissemination periods and had diffusing power only during the former period. Second, vaccine administration–related words were frequently mentioned and diffused through April to May and had diffusing power throughout the period. The background to these findings is that the sentiment of longing for vaccines outweighed that of hesitancy toward vaccines during this period.

Originality/value

This study finds that the timing of the rise in vaccine hesitation sentiment and the timing of the start of vaccine supply were misaligned. This is one of the reasons that Japan, which originally exhibited strong vaccine hesitancy, did not face vaccine hesitancy in the COVID-19 vaccine promotion project.

Citation

Tanihara, T. and Yamaguchi, S. (2024), "How was the coronavirus vaccine accepted on Twitter? A computational analysis using big data in Japan", Global Knowledge, Memory and Communication, Vol. 73 No. 8/9, pp. 1159-1173. https://doi.org/10.1108/GKMC-07-2022-0163