Exploring tourist interaction from user-generated content: Topic analysis and content analysis

This study aimed to identify the interaction actors, experiences, and results of tourists during their travels. Latent Dirichlet Allocation theme analysis was used to identify different themes in 9254 TripAdvisor items of user-generated content (UGC) and 15600 items of Sina Weibo UGC. A content analysis of 2000 UGC items was conducted using NVivo 11 software to uncover the details of tourist interactions. The study results showed that tourists preferred to share reviews on TripAdvisor and emotions on Sina Weibo. Seven types of tourist interaction actors were found: city, attractions, natural environment, companions, other tourists, residents, and service personnel. Seven emotions were generated in positive or negative interactions: joy, happiness, missing, awe, belonging, disappointment, and anger. Our findings provide insights for tourism managers and marketeers.