DataForUkraine: Adapting Social Science Tools for Crisis Response

The full-scale invasion of Ukraine by the Russian Federation shocked many scholars of the region, international observers, and policymakers. In the initial weeks and months of the invasion, relief organizations and government agencies lacked data about events on the ground and struggled to mount an effective response. The shocking extent of violence, mass flows of people, and lack of data collection infrastructure meant that there was an urgent need for new means of rapidly producing and disseminating data. At the request of policymakers at the United States Agency for International Development (USAID), a team of scholars on Ukraine and computational social scientists rapidly constructed a social media-based, near-real time event detection system to
provide publicly available data on developments in Ukraine. To accomplish this, we turned to a popular social media platform: Twitter. The popularity of Twitter among journalists, activists, and policymakers means that the platform is often the first place where events are reported and discussed. Furthermore, Twitter’s API provides easy access to rich information on the content of and engagement with tweets, providing a window into the timing and perceived importance of events across communities of users, making it an increasingly important tool for researchers. Combining deep country expertise with computational tools, the resulting system – #Data4Ukraine – distills massive volumes of information into hourly geo-coded data updated in near-real time that can both inform decision-making and answer important research questions.