HomeAbstract of 7th-ictmhWhen Did the Island Change? Machine Learning Detection of Sustainability Tipping Points in Marine Tourism

When Did the Island Change? Machine Learning Detection of Sustainability Tipping Points in Marine Tourism

  • Jun 17, 2026
  • Posted by:
  • Category: Abstract of 7th-ictmh
No Comments


Abstract Book of the 7th International Conference on Tourism Management and Hospitality

Year: 2026

[PDF]

When Did the Island Change? Machine Learning Detection of Sustainability Tipping Points in Marine Tourism

Prof Dr Narong Pleerux, Phannipha Anuraksakornkul, Thayukorn Prabamroong, Phurith Meeprom, Pasiree Yongsiri, and Parinya Nakpathom

ABSTRACT:

Understanding when sustainability issues emerge is critical for adaptive destination management, yet conventional monitoring methods often reveal problems months or years after critical thresholds have been crossed. This study addresses this challenge by applying machine learning-based change point detection to analyze temporal dynamics of sustainability issues at Koh Lan, a popular island tourism destination in Thailand. Drawing on 2,388 TripAdvisor reviews spanning 2015–2024, we developed an automated monitoring system combining sentiment analysis and statistical time-series methods to detect when tourist complaints shifted significantly. Manual content analysis of negative reviews identified 178 issue mentions across five sustainability categories: Scenery (37.64%), Staff and Service (28.65%), Facility (19.66%), Safety (10.68%), and Accessibility (3.37%). Statistical analysis detected two significant structural breaks: a 91% reduction in issues from 2018 to 2019 (p=0.031)—notably before COVID-19—and a 94% decline from 2019 to 2020 (p=0.015). Issues concentrated during peak tourism years (2016–2018, mean=50.3 issues/year) and remained suppressed through 2024 (mean=2.0 issues/year, recovering to only 11.9% of peak levels). The pre-pandemic decline suggests policy-driven improvements rather than external shocks alone shaped sustainability outcomes. For destination managers, this approach transforms user-generated content into a cost-effective, real-time early-warning system, enabling intervention at optimal moments before sustainability thresholds are irreversibly crossed.

Keywords: Destination Management, Sustainable Tourism, Real-Time Monitoring, Sentiment Analysis, User-Generated Content