Anatomy of Conflict: A Tool for Understanding Conflict Dynamics from Social Media Data

Analysis of information posted on social media has become a key application area for big data analytics. Public posts on social platforms offer an unprecedented view into the opinions, beliefs, and ideological divides of target populations. This paper describes an AI-assisted observation-based tool for analyzing belief dynamics of target populations in conflict areas based on their online posts. The tool was developed to aid peace negotiators, military/political analysts, and cultural anthropologists in exploring cultural and political differences among conflicted population segments, uncover opinion leaders on each side, discover position changes of groups over time (e.g., changes in the degree of group polarization, radicalization, or reconciliation), and understand their demographic and psychographic footprints. The paper advances a key application domain of AI-assisted big data analytics - namely, the analysis of group polarization from social media traces. Its growing importance stems from its ability to shed light on the perspectives of different sides in a conflict (as a step towards enhanced understanding), alert analysts when a situation is escalating (to encourage early remediation), and offer insights that enhance negotiation and enable mitigation.