Public perception-informed resilience management priorities for urban transportation infrastructure under extreme rainfall: evidence from social media analytics
Dedong Wang, Yongqiang Zhang, Shaoze FangPurpose
This study explores publicly perceived transportation infrastructure resilience and associated emotional responses during extreme rainfall events, with particular attention to the social dimension of infrastructure resilience.
Design/methodology/approach
Based on the robustness, redundancy, rapidity and resourcefulness (4R) resilience framework, this study analyzes social media data from seven extreme rainfall events. Bidirectional Encoder Representations for Topic Modeling (BERTopic) modeling and fine-grained emotion analysis are employed to identify resilience-related public discussion themes, examine the relative salience of different resilience attributes, and compare emotional responses across different resilience attributes.
Findings
The results show that robustness emerges as the most salient resilience attribute in the observed social media discourse. Negative emotions collectively account for a substantial share of public responses, with sadness being the most prominent negative emotion. Significant emotional differences are also observed across different resilience attributes.
Practical implications
The findings provide complementary demand-side evidence that can inform transportation infrastructure resilience governance and decision support by highlighting publicly salient resilience concerns under extreme rainfall conditions.
Originality/value
This study extends the application of the 4R resilience framework beyond a predominantly engineering-oriented perspective by incorporating public-perception analysis. By mapping social-media-derived discussion topics onto 4R and further comparing emotional-response differences across these attributes, the study provides a complementary demand-side perspective for transportation infrastructure resilience assessment.