Unexpected Damage from Typhoon KHANUN of 2023 on Izena Island: Trial of Quantitative Analysis Using Topic Model

Typhoons cause numerous casualties, extensive property damage, and significant disruptions to infrastructure. Natural disasters often result in unexpected damage, making it difficult to prepare for events beyond our imagination. It is crucial to verbalize people’s experiences during disasters and use this information in disaster prevention plans. Using computerized topic model and correspondence analysis, this study quantitatively analyzes text data from a questionnaire survey about the damage experienced on a small remote island during Typhoon No. 6 (KHANUN) in 2023. These analyses clarify what was unexpected and what equipment should be prepared based on the experience of this typhoon. Three topics revealed unexpected damage. The study also identified that emergency food should be prepared as a lesson from this typhoon. Topic model helps identify underlying patterns in the data, crucial for understanding unexpected damage. The value of this study lies in its quantitative analysis of text data and the visualization of disaster experiences. Although this research is only a small sample size contribution, it is believed that it will be useful to disaster prevention strategies in other island regions and provide methodological framework that can be applied to in the Asia–Pacific region, which often hit natural disasters.