#Coronavirus: Monitoring the Belgian Twitter Discourse on the Severe Acute Respiratory Syndrome Coronavirus 2 Pandemic

Abstract

In this study, a social media analysis is conducted to examine the public discourse about the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic on Twitter. In particular, this study aims to examine (a) how the number of tweets varies as a function of the timeline of the pandemic and associated measures and (b) how the content of these tweets, including displayed emotions, changes. Therefore, 373,908 tweets and retweets from Belgium were collected from February 25, 2020 to the March 30. Time series analysis, network bigrams, topic models, and emotional lexica were deployed for analysis. The results showed that significant events related to the virus correlated with an immediate increase in the number of tweets addressing them. Furthermore, the Belgian Twitter discourse was characterized by positively connoted words, which also refer to European solidarity. These findings do not only stress the relevance of Twitter as a medium for public discourse during lockdowns, but also seem to indicate that the Belgian public supports policy measures that respect solidarity in Europe.

What social media told us in the time of COVID-19: a scoping review

“…54 Another study analysed 373 908 Belgian tweets and retweets, which showed that the public relied on the EU coalition to tackle the pandemic. 72 De Santis and colleagues analysed 1 044 645 tweets to identify daily hot topics in Italy that were related to the COVID-19 pandemic and developed a framework for prospective research. 62 One thematic analysis study of 1 920 593 Arabic tweets in Egypt showed that negative emotions and sadness were high in tweets showing affective discussions, and the dominant themes included the outbreak of the pandemic, metaphysics responses, signs and symptoms in confirmed cases, and conspiracism.…”

Changes in Public Response Associated With Various COVID-19 Restrictions in Ontario, Canada: Observational Infoveillance Study Using Social Media Time Series Data

“…In a study of COVID-19-related tweets from February 25 to March 30, 2020, in Belgium [9], researchers plotted tweet frequency alongside major COVID-19-related events, and they found that spikes in tweet frequency coincided with COVID-19 infections, stock market crashes, school closures, and infections of notable persons. The frequency of tweets about COVID-19-related topics has also been used to measure perceived susceptibility and severity of COVID-19 and was correlated with interventions, public events, and case counts [10].…”

“…The limited number of studies that examined the association between COVID-19-related events, restrictions, and public opinion have typically approached the question from a descriptive manner, such as by graphically plotting major events and COVID-19 incidence on a timeline against COVID-19-related tweet frequency [9]. However, without considering the contribution of multiple factors simultaneously (business closures, school closures, holidays, other restrictions, etc), such as through the use of multivariate time series analysis, these studies may over-or understate the unique contribution of any given factor due to statistical confounding.…”

Analysing discourse around COVID-19 in the Australian Twittersphere: A real-time corpus-based analysis

“…Finally, in a study of public discourse about the pandemic in the Belgium Twittersphere, Kurten and Beullens (2020) conduct a time-series analysis of tweets containing a selection of pandemic-related hashtags collected through Twitter's 'statuses/filter' streaming API endpoint. They used this collection to generate both topic models and to analyse changes in sentiment over time.…”

“…Due to the large volume, velocity, and variety of data about the pandemic available through Twitter, it has been treated as a useful source of big data (Brooker et al, 2016;Felt, 2016;Kitchin, 2014;Kitchin & McArdle, 2016). Researchers have utilized data sourced from Twitter to examine various aspects of the ongoing pandemic (Bowe et al, 2020;Gruzd and Mai, 2020), including a growing body of work that examines public discourse about COVID-19 (e.g., Damiano and Catellier, 2020;Das and Dutta, 2020;Jimenez-Sotomayor and Gomez-Moreno, 2020;Kurten and Beullens, 2020;Lwin et al, 2020;Wicke and Bolognesi, 2020;Xue et al, 2020). It is clear from these studies that public discourse about the pandemic appearing on Twitter and other social media platforms consists of an interconnected set of more or less cohesive topics that are directly or indirectly linked to COVID-19.…”