Microblog data: Measuring tourism destination image using importance-performance analysis

Abstract

Understanding the perceived image of a tourism destination is crucial for destination management organizations to effectively align their strategic efforts in shaping and promoting tourism attributes. This study presents a fully automated text mining approach to empirically measure the perceived image of tourism destinations as represented on X, illustrated by means of four Brazilian coastal tourism destinations. First, a classification algorithm determines importance values for each of the 14 types of tourism defined by the United Nations World Tourism Organization (UN Tourism). Second, a multilingual XML-RoBERTa model optimized for sentiment detection from tweets assesses each destination’s performance across each tourism type. Third, the importance and performance measures are combined in grids to depict the perceived image of the destinations. This fully automated text mining technique represents a ready-made solution which enables comparability between any two or more destinations through its comprehensive attribute-based approach.