EU-border crisis on Twitter: sentiments and misinformation analysis

The objective of this paper is to investigate the information and to detect the presence of misinformation on Twitter posts circulating in relation to migration events happened in 2020 at the Greek-Turkish border and in 2021 at the Polish-Belarusian border. Data were retrieved through API by using keywords referring to the two border events. The study was carried out by applying text mining and sentiment analysis techniques on tweets and retweets related to these two events, and by conducting a qualitative analysis on specific subsets of tweets. Our results show that in both borders’ crises migration is perceived as an emergency issue, migration-related narratives mainly refer to “war”, “attacks”, “tension”, “invasion” and the emotions expressed are mostly negative. In addition, in outbreaking crisis, the identification of misinformation in social media is extremely challenging, because of the rapid circulation of rumours related to facts that are rather difficult to ascertain.