A POSTECH study using AI and NLP revealed that different types of weather forecast errors, like overestimation or underestimation, trigger distinct emotional responses in the public during natural disasters. The research examined reactions to Typhoon Khanun.