Social Computing for Post-Earthquake Vulnerability and Resilience Assessment

Social media offers significant potential to transform disaster communication and, in turn, enhance community resilience to seismic events. This study integrates social media sentiment with hazard data, damage assessments, and social vulnerability indices to address critical gaps in disaster resilience research. While prior work has highlighted the value of social media for real-time insights, few studies have effectively correlated sentiment with hazard exposure, physical impacts, and social equity considerations. Focusing on the Ridgecrest earthquake sequence of July 2019, the analysis captures both temporal and spatial patterns of community response. The findings reveal a decline in positive sentiment and a rise in negative sentiment during the earthquakes, reflecting heightened concern and anxiety. Sentiment analysis also provides clearer insight into hazard perception than traditional measures, showing weak correlations with peak ground acceleration and peak ground velocity but stronger alignment with subjective Did You Feel It? reports. Finally, integrating the Social Vulnerability Index with sentiment data demonstrates a promising approach for identifying at-risk communities and guiding resource allocation. Together, these results underscore the value of incorporating social media analytics into disaster response frameworks to strengthen community resilience.