External-Facing Communication on Social Media During the Russia-Ukraine Conflict of 2014

Paper Title

External-Facing Communication on Social Media During the Russia-Ukraine Conflict of 2014

Publication Info

• Topic area: Social media communication in geopolitical conflicts
• Keywords: Russia-Ukraine conflict, external-facing communication, social movements, Twitter analysis, hashtag projection, political alignment, topic modeling, inauthentic activity, crisis informatics, social media activism

Background and Problem

• Problem / challenge: Limited research exists on how social movements in complex geopolitical contexts communicate with external audiences, especially during prolonged crises.
• Significance: Understanding external-facing communication is critical for social movements to mobilize global support and resources, particularly when local efforts are insufficient due to adversarial external actors.
• Motivation and related work: Previous studies have explored social media use during crises, adversarial messaging, and misinformation, but have not specifically examined external-facing communication in movements competing for global attention during geopolitical conflicts. This paper addresses that gap by analyzing the Russia-Ukraine conflict of 2014.

Solution

• Proposed approach: Analysis of external-facing communication on Twitter during the 2014 Russia-Ukraine conflict using topic modeling, hashtag projection, and qualitative coding.
• Novelty:
  1. Characterization of external-facing communication content and practices of opposing social movements within the same geopolitical context.
  2. Identification of challenges in engaging external audiences, including inauthentic activity and hashtag hijacking.
  3. Development of coding schemes and metrics for analyzing topics, hashtags, domains, and participation in prolonged conflicts.
  4. Insights into strategies for sustaining long-term participation and engagement in social movements.
• Procedure and key techniques:
  • Data collection of 58M tweets over 21 months, filtered to 823,793 relevant English tweets.
  • Categorization into pro-Ukrainian, pro-separatist, and neutral discourses using hashtag projection and political alignment scores.
  • Topic modeling with BERTopic and qualitative coding to identify 21 themes.
  • Analysis of hashtags, geographic references, linked URLs, user roles, and persistence to compare communication content and practices across discourses.

Results

• Concrete findings:
  • Pro-Ukrainian discourse focused on international law and justice, using specific hashtags like #russiainvadedukraine, but struggled to sustain long-term user participation.
  • Pro-separatist discourse employed generic hashtags and first-person video content, with evidence of inauthentic activity from users in India posing as Western individuals.
  • Neutral discourse featured broader engagement with firsthand accounts and diverse news sources.
• Advantage over baselines:
  • The study provides a nuanced understanding of external-facing communication strategies and challenges, highlighting differences in content focus, dissemination practices, and user engagement across opposing movements.
• Experiments / evaluation:
  • Data processing pipeline included language filtering, drift removal, topic modeling, and political alignment scoring.
  • Evaluation of themes, hashtags, geographic references, and URLs to compare discourses.
  • Analysis of user roles (producers vs. diffusers) and persistence levels to assess engagement.
• Limitations and future work:
  • Limited ability to verify the authenticity of user activity and content linked via URL shorteners.
  • Future research could explore coalition-building among social movements, design interventions for targeted engagement, and the role of identity in connective action.

Summary

This paper investigates external-facing communication on social media during the 2014 Russia-Ukraine conflict, analyzing pro-Ukrainian, pro-separatist, and neutral discourses. It identifies differences in content focus, dissemination practices, and user engagement, revealing challenges such as inauthentic activity and hashtag hijacking. The findings suggest that social movements can benefit from strategies emphasizing eyewitness accounts, targeted engagement, and coherent identity development. The study contributes coding schemes, metrics, and insights that can inform future research and design interventions to support social movements in achieving their goals.