Digital resilience of sports communities in a Facebook group during military conflict

Background and Study Aim. In the context of military conflict, analyzing online interactions becomes a crucial tool for studying social and emotional processes. This study aims to analyze the sentiments, activity patterns, and key themes of interaction among participants in a sports Facebook group during wartime. Materials and Methods. The study was based on the analysis of data from 1,055 posts by members of a Facebook group. Sentiment analysis, topic modeling (Latent Dirichlet Allocation, LDA), and clustering (KMeans) methods were employed. The data included text messages, the number of likes and comments, as well as user activity metrics. Word clouds and correlation heatmaps were used to visualize key aspects. Results. The analysis revealed a significant proportion of neutral messages (approximately 90%), reflecting the nature of discussions within the group. The emotional tone of the posts was predominantly neutral, with a small number of positive and negative messages. Clustering of group participants identified four main activity types: low, medium, high, and anomalously high. Additionally, topic modeling identified key themes related to physical activity, social interactions, and the influence of current events. Conclusions. The findings highlight the importance of analyzing online communities to understand social and emotional processes in crisis situations. The key discussion topics and emotional tone can be valuable for developing support strategies for participants and improving engagement in online groups.