Mapping sentiments and emotions in a UNESCO World heritage destination

The rapid development of Internet-based applications, including websites, social networks and blogs, has led to a great number of opinions and reviews being generated about products, services and activities. In this study, an analysis of the emotions and feelings found in reviews of the ten main attractions in Coimbra on the TripAdvisor site is presented. A total of 436 reviews was collected, dating from the years 2022 and 2023. The VADER model (Valence Aware Dictionary and Sentiment Reasoner) was used for sentiment analysis, allowing sentiments to be classified as positive, negative or neutral. The confusion matrix showed the VADER model to be highly accurate in detecting positive sentiments but that it has a lower level of performance for neutral and negative sentiments. The results of this analysis show that 88.30% of reviews are positive, 9.40% negative and 2.30% neutral. The trained model ‘j-hartmann/emotion-english-distilroberta-base’ was used for analysing emotions, and identified Ekman’s six basic emotions, with enjoyment standing out as predominant (46%). Negative emotions, although less frequent, represent 14% of the total. This study also reveals a high degree of positive feelings towards the attractions of a city like Coimbra, a UNESCO World Heritage Site, with a prominence of enjoyment. These results can provide destination management organisations (DMOs) and stakeholders or destination management companies (DMCs) with a more thorough understanding of visitor experiences and perceptions, enabling them to improve services and target marketing strategies more effectively.