CEU Electronic Theses and Dissertations, 2025
| Author | Kabdula, Asset |
|---|---|
| Title | Insights into Visitor Perceptions of Kazakhstan's Tourism Sector: Sentiment Analysis and Topic Modeling of TripAdvisor Reviews |
| Summary | Kazakhstan’s tourism industry has experienced rapid growth in recent years, yet academic re- search on tourists’ perceptions and satisfaction in this context remains limited. This study lever- ages user-generated content from TripAdvisor to provide a comprehensive sentiment and topic analysis of visitor experiences across Kazakhstan’s attractions from 2011 to 2024. Employing advanced machine learning and natural language processing techniques including BERT-based sentiment classification, Non-negative Matrix Factorization (NMF) for topic modeling, and zero-shot topic classification over 23,000 multilingual reviews covering 1,001 attractions were systematically analyzed. The results reveal a strong positive correlation between TripAdvisor review activity and official international tourist arrivals, supporting the use of digital trace data as a proxy for real- world tourism trends. Sentiment analysis indicates an overall upward trend in positive senti- ment among tourists, particularly after 2020, coinciding with the State Program for the Devel- opment of the Tourism Industry (2019–2025). Topic modeling identifies seven key themes such as transportation, tourism services, culture and history, nature and hiking, religious and spiritual sites, urban parks, and food and leisure with transportation-related reviews consistently asso- ciated with lower sentiment, highlighting persistent challenges in infrastructure. Conversely, nature and hiking topics exhibit the highest sentiment scores, emphasizing Kazakhstan’s natural attractions as a key strength. The study further finds no local bias in online review data and finds no significant differ- ence in sentiment between reviews submitted by Kazakhstanis and those provided by foreign visitors. Additionally, there is a statistically significant seasonality in sentiment, with peaks in November and troughs in March. These findings offer actionable insights for policymakers, destination marketers, and tourism authorities, underscoring the value of digital review ana- lytics in enhancing tourism strategies and service quality. By filling a significant gap in the literature, this research provides a robust, data-driven understanding of tourist perceptions and evolving trends in Kazakhstan’s tourism sector. |
| Supervisor | Posfai, Marton |
| Department | Network Science MSc |
| Full text | https://www.etd.ceu.edu/2025/kabdula_asset.pdf |
Visit the CEU Library.
© 2007-2025, Central European University