Identifying unique features of the image of selected cities based on reviews by TripAdvisor portal users

Abstract

ABSTRACT The paper aims to identify unique features in the image of four Baltic cities: Gdansk, Kaliningrad, Riga, and Szczecin, based on an analysis of reviews posted on the TripAdvisor portal. The text mining technique was used to extract the words most frequently used in opinions, while sentiment analysis was performed to assess the strength of negative and positive reviews. Analysis of variance was used to extract the unique and common features of the image of each city analysed. The results showed that Riga and Gdansk have the largest number of unique features/attributes, while Kaliningrad has the smallest. Positive and negative sentiment analysis indicated that Gdansk and Szczecin have a higher proportion of positive sentiment in reviews than Riga and Kaliningrad. The study confirmed the importance of traveller-generated content as an image-building agent, and shows that destination image attributes can be effectively identified using text mining in both the cognitive and affective dimensions. It also showed that it is possible to identify significant differences in the image of a destination, which can subsequently be used by DMOs in the branding process to distinguish destinations from one another on the tourism market.