TOPIC MODELLING IN TWEETS ON RUSSIA’S INVASION OF UKRAINE
Maria Ioana Popa. Faculty of Economics and Business Administration, University of Craiova. E-mail: popa.maria.w3z@student.ucv.ro
Ioana-Andreea Gîfu. Faculty of Economics and Business Administration, University of Craiova. E-mail: gifu.ioana.w3b@student.ucv.ro
Vasile Georgescu. Doctoral School of Economics, University of Craiova. E-mail: v_geo@yahoo.com
- Fuzzy Economic Review: Volume 28, Number 1, 2023 (On Going)
- DOI: 10.25102/fer.2023.01.03
Abstract
The Ukraine–Russia crisis erupted into a war on February 24, 2022, due to Russian invasion of Ukrainian territory, and quickly became one of the hottest topics on Twitter (now rebranded as X), with millions of tweets generated every day. In this paper, we use Natural Language Processing methods, such as word cloud techniques and topic modelling, to analyze people opinions and their reactions on various events related to the war. Our approach consists of splitting this huge number of tweets in different corpora by means of convenient keywords (for example, “Bakhmut”, “Zaporizhzhia") and to extract the most relevant information from each corpus, via topic modelling, with the aim of highlighting major themes of interest and thus leading to a better understanding of public thinking and attitude about Russia's aggression against Ukraine.