An exploratory study on the relationship between regional tourism issues and the Regional Extinction Index (REI): Applying LDA-LSTM to the case of Haeundae district, Busan

This study explores the relationship between regional tourism issues and the Regional Extinction Index (REI), focusing on Haeundae-gu, Busan. For this purpose, news articles spanning 15 years (2010–2024) were collected and analyzed using Latent Dirichlet Allocation (LDA) to identify major tourism topics, and Long Short-Term Memory (LSTM) modeling was then applied to predict the temporal dynamics of these topics, followed by a lagged correlation analysis between the predicted topic distributions and the REI. The results indicate that tourism-related news encompassed diverse themes, with specific issues showing significant correlations with the REI at varying time lags. Specifically, development-oriented tourism issues, such as infrastructure expansion and urban planning, exhibited a negative correlations with the REI, whereas community- and environment-oriented issues demonstrated a positive correlations. These findings suggest that the demographic impacts of tourism issues vary according to their nature and may manifest with temporal delays. This study highlights the necessity of interpreting tourism not merely as an economic driver but as a potential long-term factor influencing demographic structural changes.