Analysis of Dunhuang tourism hotspot topics based on the LDA model

Abstract

Dunhuang, as a significant cultural hub along China's Silk Road, attracts numerous domestic and international tourists with its rich cultural heritage and unique natural landscapes. However, balancing the preservation of cultural heritage with the demands of tourism development, while comprehensively enhancing the quality of tourist experiences, has become a critical focus of current research. This study analyzes 5,000 valid online reviews from two major online travel platforms, Ctrip and Dianping, collected between January 2020 and October 2024. Using the Latent Dirichlet Allocation (LDA) model, the research conducts topic modeling and sentiment analysis of tourist reviews to identify the core attractions and potential issues in Dunhuang tourism. The findings reveal that cultural heritage sites, such as the Mogao Caves and their murals, and natural attractions, including Mingsha Mountain and Crescent Lake, serve as the primary draws for tourists. Key factors influencing tourist satisfaction include service quality, management effectiveness, and infrastructure convenience. In response to tourist feedback on issues such as peak-season management, insufficient tour guide services, and infrastructure limitations, this study proposes specific strategies, including optimizing reservation systems, promoting intelligent management, and leveraging digital technology in cultural dissemination. The study innovatively develops a multidimensional tourist experience analysis framework, expands the application of text mining techniques in cultural tourism research, and provides theoretical support and practical guidance for the high-quality development of Dunhuang tourism and the digital transformation of cultural dissemination.