Text-based modeling reveals the sector-specific benefits of emerging technologies for extreme flood adaptation

Technological progress can help reduce the flood adaptation gap, but its sector-specific impacts remain unclear. Here, we developed a text-mining approach to analyze how four technologies—drones, online rescue forms, navigation apps, and drainage systems—affect perceived flood losses. We applied this model to 3.58 million social media posts from extreme floods in two Chinese cities. Results show highly heterogeneous effects: drones and online rescue forms support multisectoral recovery, while navigation apps and drainage mainly target traffic and buildings. For reducing casualties and restoring communication and water/electricity supply, drones (43.0%–62.2%) and online rescue forms (19.5%–37.1%) ranked higher in importance than rescue effort effectiveness (17.3%–28.5%). Technology adoption reduced perceived losses primarily by improving rescue effectiveness, with online rescue forms showing the strongest propagation path (coefficient: 0.577–0.579). This study provides a text-based framework for evaluating technology in flood adaptation and supports planning under resource constraints.