Gender variation in the discourse of online hotel reviews: A corpus-based semantic analysis

The study deals with sematic categories mentioned in a 24,6-million-word corpus of online hotel reviews from Booking.com authored by men and women. The overall objective of the current study is the application of text mining techniques, namely corpus-based semantic analysis, to find if there are significant differences in the discourse of men and women when reviewing hotels. The reviews employed for the analysis were written by people located in the UK and relate to hotels and apartments of all star ratings. The semantic categories suggested by McArthur (1981) were detected by means of sematic tagging with UCREL Semantic Analysis System (USAS). We found a number of statistically significant differences between male and female discourses both in terms of linguistic expression and highlighted aspects of the stay. For example, women's writing appears to be more personal and emotional with an extensive use of pronouns and boosters. As for the content of reviews, males tend to mention money paid, the location of the hotel and food; while females are more likely to write about family and friends or partners.