Russia-Ukraine War: A Topic Modeling and Sentiment Analysis Study based on X (formerly Twitter) Data
Keywords:
Topic Modelling, Sentiment analysis, Crisis communication, Social media AnalyticsAbstract
Purpose: The purpose of this study is to explore and analyse the discussed topics and sentiments expressed by individuals in their tweets related to the Russia-Ukraine War.
Methodology: The Tweet Archiver was used to retrieve the data regarding the Russia-Ukraine war from August 30 to September 30, 2022. During this specified timeframe, a total of 49994 Tweets were retrieved for analysis. A comprehensive analysis of data was performed by employing Python and its libraries. Topic analysis was carried out using BERTopic. The VADER library was used for sentiment analysis.
Findings: The study found ten predominant topics around which the discussions revolved. The most prevalent topic with the highest number of tweets is “NATO’s Involvement in Russia-Ukraine Conflict”, followed by “Geopolitical Tensions and Energy Politics in China, India, Russia, and Ukraine”. Moreover, the sentiment analysis demonstrated that the majority of tweets reflect negative sentiments, while the least emulate positive sentiments. However, a good number of tweets assumed a neutral standpoint towards the war. The geographical distribution of tweets highlights that tweets were poured in from 124 countries across the globe, with a significant contribution from the USA.
Originality: The study illustrates the sentiments and the topics of discussion regarding the Russia-Ukraine war, gains some insights into overall public communication about the war, and complements the existing literature.