Analyzing Replies and Interactions Among Users with Different Stances: A Case Study of the Russia-Ukraine Conflict
摘要
Since its outbreak, the Russia-Ukraine conflict has been extensively discussed on social media. Analyzing the replies and interactions across different stances is significant for comprehending the political, economic, military, and diplomatic factors, as well as the increasing polarization that has been widely observed among online discussions. This paper analyzes the replies from Reddit discussions, focusing on user replies to understand sentiment, topics, and interaction patterns among users with Pro-Russia, Neutral, or Pro-Ukraine stances. Descriptive statistical analysis is employed to study word frequency, entity frequency, and emotion distribution. Topic modeling is utilized to extract topics from replies with different stances, revealing differences in the content and providing insights into stances formation. User interaction analysis indicates that replies between opposing stances tend to exhibit stronger negative sentiment, and different stances also reduce the likelihood of interaction. Pro-Ukraine users prefer interacting with like-minded users, while neutral users interact more equally with different stances, aiming to ease tensions and foster discussion.