Evaluating Tourist Dissatisfaction with Aspect-Based Sentiment Analysis Using Social Media Data

Abstract

Tourism satisfaction is essential for encouraging tourists to stay longer, spend more and return. However, visitor dissatisfaction can also prove useful for understanding any shortcomings of a tourist destination, and Twitter, Instagram and TripAdvisor reviews might be able to provide an insight into tourist perceptions and experiences. This study examines the major causes of tourist dissatisfaction with a tourism destination using an aspect-based sentiment analysis approach to understand the key points of negative tweets, posts or reviews. We examined 19,340 tweets, 7,712 Instagram posts and 25,483 reviews about Granada in Spain in order to evaluate the negative user's perceptions, discover management-related problems and provide feedback to destination management organizations to enable them to improve their services and operations. Our work contributes to computational methods to address tourism (dis)satisfaction with a process to identify the most important entities (places), an algorithm to identify aspects and opinions, and the use of word-trees to show the most important aspect-opinion tuples. In practical terms, we provide to tourism industry professionals and managers, as well as travelers, with methods to identify the reasons for tourist dissatisfaction from available social media data, in such a way that managerial strategies or travel plans can be improved.