Tell me about your culture, to predict your tourism activity preferences and evaluations: cross-country evidence based on user-generated content

ABSTRACT This study extracts user-generated content from Tripadvisor to understand the relation between national culture with tourists’ preferences and evaluations of activities at destinations. Travel reviews were analyzed to measure the tourists’ preferences for three types of tourism activities using machine learning methods. Multiple regression analysis was used to investigate the relations between Hofstede's cultural dimensions, tourists’ preference and evaluation of tourism activities. Findings show that national culture significantly relates to tourists’ preferences for only one type of activity (i.e. participatory), while it does significantly relate to tourists’ evaluations of all three types of tourism activity (sightseeing, participatory, and ancillary activity). Tourists’ cultural dimensions are found to have differential relations with tourists’ evaluations of tourism activities. Power distance, individualism, and masculinity have positive relationships with tourist evaluation. Conversely, uncertainty avoidance, long-term orientation, and indulgence are negatively related to tourist evaluation. The study concludes by discussing the theoretical and practical implications.