Sentiment Analysis of Tweets on the Russia-Ukraine War

This study aims to mine public opinion on the X platform using sentiment analysis concerning the Russia-Ukraine war. The study seeks to address the contradiction in recent reports regarding public sentiment toward the conflict between Russia and Ukraine. The analysis used both supervised and unsupervised learning approaches. Under supervised learning, both BOW and TF-IDF features were used through CountVectorizer together with logistic regression and support vector machine to predict sentiments. TextBlob, SentiWordNet, AFINN, and VADER were used under unsupervised learning to measure subjectivity and polarity in the text data. The study found that public opinion supports Russia resulting in the majority of pro-Russia tweets. Even though public opinion about Russian leader, Putin is negative, his name is the most trending name on the X platform regarding this topic. On the other hand, even though tweets are slightly skewed towards supporting Russia the general sentiment on war is negative. The outcome of this analysis could significantly impact decision-making in political conflicts in general. The findings highlighted the impact of leader-centric influence on social media platforms in shaping public opinion. The results also showed that influential figures or leaders can strategically utilize social media to amplify their narratives during a conflict situation, though it contradicts public opinion. The study contributes by addressing the contradictions in recent reports and providing an understanding of public sentiment towards the Russia-Ukraine war on the X platform. It highlighted the complexity of public opinions by differentiating tweets between pro-Russia and pro-Ukraine sentiments regarding the war addressing few limitations in prior studies.