Rethinking the community detection for tourist mobility networks: special structures and their implications for tourism

Recently, there has been growing interest in the community structure within tourist mobility networks. However, existing studies are often based on vague or overly simplistic assumptions, such as presuming communities as mutually disjoint and static entities, which may fail to capture the complexity of tourist behaviour and potentially lead to misinterpretations of the tourism space. In this paper, we argue that community detection for tourist mobility networks demands more nuanced and rigorous considerations. We introduce four special types of community structures—overlapping, multiscale, dynamic, and multilayer—and present four corresponding empirical studies to examine each, using a social media dataset of tourist activities in Suzhou, China. The findings demonstrate that different assumptions in community detection for tourist mobility networks can lead to notably different spatial interpretations, underscoring the need for context-specific and flexible methodologies. Overall, this paper conceptualises the tourism space under mobility as an emergent and networked system—one shaped by overlapping, multiscale, dynamic, and multilayer interactions rather than fixed boundaries. It also provides methodological guidance for analysing this complexity and demonstrates how such structural characteristics can inform adaptive destination planning and management.