Cruise shipping network analysis

Cruise shipping is one of the fastest-growing sectors in the maritime industry. It provides unique services for passengers both onboard the ship and at cruise destinations around the world. Cruise shipping networks consist of ports as the nodes and cruise sailing routes as the links. As demand and competition in the sector grow, cruise lines need to expand their networks by adding new routes and destinations. Cruise networks also change in response to events such as the outbreak of the COVID-19 pandemic or the need to focus on new destinations. As a result of such changes, some ports become better connected than others. Network analysis has become an important research tool applied to various disciplines, such as sociology, biology, computer science, economics, international trade, information technology, geography and transport. However, analysis of maritime networks, especially for the cruise shipping sector, is limited. This thesis aims to analyse the Asian-Australasian cruise shipping network comprising cruise ports as nodes (vertices) and cruise shipping routes as links (edges) that connect them. The analysis covers four main aspects of the network, namely formation, connectivity, impact analysis and simulation, to gain insights into the network. First, the study seeks to analyse the development of the network using the Erdös-Renyi model and the exponential random graph model (ERGM). Second, the study analyses the effect of port attributes such as the number of attractions, travellers' ratings and geographical locations on network connectivity using the exponential random graph model and network association measures, e.g. assortativity and modularity. Third, the study uses various network measures to analyse the changes in the network structure, characteristics and connectivity caused by the COVID-19 pandemic. Fourth, using network simulation based on the Erdös-Renyi and exponential random graph models, the study shows how the network may respond to a change in demand for cruise service. An analysis has found that the network has a scale-free property with a degree distribution following the power law. Singapore, Shanghai and Hong Kong are hub ports controlling network connectivity. The connectivity is influenced by ports' geographical location and rating scores. The network consists of six clusters, most of which are led by hub ports. Ports are densely connected to other ports in the same cluster but sparsely connected to ports in different clusters. It has also been found that the network structure and properties have changed under the impact of the COVID-19 pandemic. Interestingly, the network becomes better connected with a larger number of links while remaining scale-free. At the same time, it has a higher connectivity efficiency with a smaller average path length and a higher clustering coefficient. The hub ports still maintain their important role with an extensive connection with other ports in the region. However, their importance has varied. The role of some ports has improved with more links, while that of others has diminished with fewer links. The result of the post-pandemic network simulation analysis shows that those ports with a larger number of triangle relationships have a better chance of getting new links, while ports with no or few connections have a minimal chance of acquiring new links. The study offers an insight into the cruise sector, especially its structure and connectivity. It also provides useful recommendations for cruise shipping and port management on how to evaluate and improve cruise lines and ports' strategic role and connectivity in the network and manage the effect caused by major events such as a global pandemic.