Social Media Migration Tracking Dataset: Twitter-based Global Discourse Analysis 2017-2020

Over the last decade, social media has emerged as a valuable tool for reflecting societal dynamics. As people spend more time online, their online traces become a window into their behaviour, choices, curiosity, whereabouts, and opinions. This study leverages a dataset of 486 million messages collected from Twitter (now X) between 2017 and 2020 to explore the potential of social media platforms as a mirror for migration trends, as reflected in the posts of users. We examine correlations between migration-related discussions on Twitter and official United Nations (UN) data, identifying significant alignments that validate the platform's utility in tracking migration dynamics. Through event detection, we highlight the responsiveness of social media to key migration events, demonstrating its value as a real-time sensor of societal reactions. Additionally, we investigate the impact of the COVID-19 pandemic on migration-related discourse, revealing shifts in public sentiment and focus during this global crisis. Our findings underscore the significant role of social media in complementing traditional data sources, providing timely insights into migration trends and public perceptions, and, in the case of COVID-19, the impact of exogenous shocks on online discourse.