Mapping Visitors Experiences—A Case Study on Selected Airbnb Datasets in Southern Europe

The study of visitor experiences incorporating user-generated content from accommodation platforms is a distinct trend in the geography of tourism. Our research aims to better understand how this content can be instrumentalized in order to assess the viability of Airbnb accommodation ranks based on specific case studies from Greece, Italy, and Spain. The methodological frame we propose is based on two separate tools. First, the users’ reviews are summarized using sentiment analysis techniques, and the positive component is extracted as a separate indicator (predictor). The second step consists of mobilizing geographically weighted regression (GWR) to identify of the potential statistical association between the Airbnb apartments’ ranks and the quantitative outputs of sentiment analysis. The results obtained for 13 case studies are based on more than 4.6 million reviews. They clearly emphasize a gap between the rank proposed by the platform and the positive scores of sentiment analysis for the accommodation units analyzed (88,053). Despite some limitations linked to the quantity of data needed to be integrated into the investigation, the methodological frame can be transferred to other destinations, providing useful information about the potential distortions of tourism markets by the meta-description of accommodation systems.