Finding patterns in urban tourist behaviour: a social network analysis approach based on TripAdvisor reviews

Abstract

Developments in ICT and the massive growth in social media usage have increased the availability of data on travel behaviour. This brings an array of new possibilities to improve destination management through Data-driven decisions. This data, however, needs to be analysed and interpreted in order to be beneficial for destination management. Different kinds of methodologies and data have already been applied to analyse spatial behaviour of tourists between and within destinations. The novelty of our paper in this sense that we apply a relational approach by conducting a network analysis methodology on a readily available big data source: user generated content (UGC) from TripAdvisor. The collected data from the city of Antwerp, Belgium shows how locals, Belgians, Europeans and non-Europeans have distinct review patterns, but also shows recurring behavioural patterns. By comparing the relational constellation of the review network to the spatial distribution of central and peripheral attractions, hotels and restaurants, we discuss the added value of social network analysis on UGC for translating (big) data into applicable information and knowledge. The results show a dominant position of a limited number of clustered attractions in the historic city centre, and shows how geographical proximity and relational proximity are interrelated for international reviewers but less for domestic reviewers. This finding is translated into a set of recommendations for policy makers and destination managers trying to accomplish a better distribution of tourists over the entire destination.