Detection and analysis of overlapping communities in the global shipping network

The global shipping network is a vital infrastructure for international trade and plays a key role in facilitating connectivity and functionality through its community structures. This study develops the Improved Speaker-listener Label Propagation Algorithm (ISLPA) to identify overlapping communities in the 2019 global shipping network, with a focus on the port distribution, hierarchical structure, and node characteristics of overlapping ports in communities of different sizes. The results indicate that: (1) Significant overlap exists among communities within the global shipping network, with communities of various sizes exhibiting dual characteristics of global integration and regional clustering. Especially, overlapping ports between large and medium-scale communities function as bridges and intermediaries in the global shipping network, enhancing connectivity across multiple regions. (2) Port nodes within large and medium-scale communities can be categorized into core, intermediate, and edge layers, with overlapping ports mainly distributed in the intermediate and edge layers. (3) Overlapping ports exhibit multidimensional structural characteristics within the shipping network, with degree centrality revealing the connectivity of ports within local networks, betweenness centrality highlighting the role of ports as network intermediaries, and node importance signifying their position in the overall network's stability and influence. The results suggest that ports with moderate degree centrality or high betweenness centrality do not necessarily function as overlapping ports. This study offers a novel perspective for understanding overlapping community structure and hierarchical characteristics of the global shipping network, providing theoretical support for analyzing the multiple functions of ports within global supply chains.