A Text Mining-Based Analysis of Online Perception Shifts in Air Travel Before and After a Major Aviation Accident

This study utilizes text mining techniques to quantitatively analyze how public perceptions and emotions toward air travel changed following the J Airline disaster on December 29, 2024. Data were collected from Naver, Daum, and Google over two six-month periods before and after the incident. Analytical methods included word frequency analysis, TF-IDF analysis, semantic network analysis, CONCOR analysis, and sentiment analysis. The results showed that overall interest in air travel remained relatively stable despite the incident. However, the focus of online discourse shifted from price-related topics to safety concerns. Keywords such as “time” and “portable battery” gained prominence in the post-disaster period. The emergence of a new cluster labeled “in-flight risk factors” in the semantic network analysis reflected growing public concern about cabin safety. Sentiment analysis revealed a 13.7-point increase in negative sentiment and more frequent use of fear-related terms such as “fear,” “fright,” and “pain.” These findings suggest that the disaster significantly heightened emotional responses related to air travel. The study provides theoretical insights and practical implications for aviation risk communication and policy design.