Figure 4
From: Early warnings of COVID-19 outbreaks across Europe from social media
source software QGIS 3.10 LTS (https://qgis.org/it/site/). (b) Scatter plot of the relationship between population size of European regions and number of unique users discussing Coronavirus in the same 2019–2020 winter season on a log–log scale. Coefficient of determination R2 for the linear regression model shows a goodness of fit equal to 0.968. Data on population size was obtained from the official COVID-19 data set (source: John Hopkins University; https://github.com/CSSEGISandData/COVID-19).
Geo-localization of tweets concerned with Coronavirus posted across Europe and relationship between number of users and population size. (a) Geographic distribution of unique users discussing Coronavirus in the winter season between 1 December 2019 and 30 January 2020. The map was generated using the open-
