Through an analysis of 20,573 TripAdvisor reviews of Vietnam’s Cu Chi Tunnels, guided by the Experience Economy Model and employing advanced machine learning techniques for topic modeling and aspect-based sentiment analysis, this study uncovers significant educational, entertainment, escapist, and aesthetic visitor experiences. It significantly advances tourism and marketing literature by demonstrating the efficacy of these analytical methods in dissecting the complex visitor perceptions at historical tourism sites. The findings not only offer pragmatic insights for site management and marketing strategies aimed at enhancing visitor engagement but also suggest a scalable framework for applying similar approaches to other dark tourism sites, emphasizing the pivotal role of curating historical narratives that align with contemporary visitor expectations.