Transcending Hotel Star Ratings: A Comparison of Topic Modeling and Sentiment Stratification in Cross-Cultural Reviews

This study examines online reviews from Chinese and international tourists staying at four-star hotels in Qingdao through the lens of Cross-Cultural Theory and the SERVQUAL framework. Drawing on natural language processing methods, including LDA topic modeling and sentiment analysis, the research identifies key themes related to facility configuration, cost-effectiveness, interactive experiences, and proximity to attractions. The results reveal that Chinese tourists emphasize family-oriented amenities and value-added services, while international tourists highlight the importance of personalized and efficient service delivery. Sentiment variations also emerge across different travel types and provincial backgrounds, indicating differentiated expectations within tourist groups. Based on these findings, the study suggests tailored service strategies to meet the diverse needs of culturally heterogeneous customers and underscores the role of intelligent technologies and effective language communication in enhancing satisfaction and competitiveness in the four-star hotel segment.