Using cellular data to analyze the tourists' trajectories for tourism destination attributes: A case study in Hualien, Taiwan

Abstract

This study adopts social network analysis (SNA) to identity the spatial role of nodes in a tourism region. The trajectory of multi-destination tourists will provide great information for tourism management, but tracking tourists' movements is a big challenge. Thanks to the popularity of mobile communication devices (mobile phones) nowadays, these can passively or actively provide information on owners' movement. This study uses call detail records (CDR), which are mobile phone positioning data, to construct a trajectory network of tourists in Hualien, Taiwan. Three types of centralities, as well as brokerage analysis and role analysis, are used to measure the relations in a complicated network. Degree centrality (DC) reveals popular nodes that have good relationships with others directly. Reach closeness centrality (RCC) shows indirect relationships with others. Betweenness centrality (BC) indicates the most important mediator nodes that help others to connect. BC and brokerage analysis are used to classify destinations into five brokers: the coordinator, consultant, gatekeeper, representative, and liaison. We also identify social positions of each destination by role analysis. This study identifies a total of 78 nodes (origins and destinations) and their functions in the tourism network. We find some nodes have more than one mission, while some only have one. The functions of nodes can be different on weekdays and weekends. We distinguish the tourism region into different size of districts. The results of this study can help managers consolidate strategies for tourism marketing.