Multi-feature fusion for the evaluation of strategic nodes and regional importance in maritime networks

Abstract

Node importance has been a widespread research topic owing to the impact of uncertainties and accidents on supply chains during maritime transport. Although the analysis and investigation of critical nodes using complex network theory is mature and systematic, there is often a lack of multiscale node identification models and theoretical frameworks. This paper proposes a novel quantitative analysis framework and process for node importance by fusing multiple features. Node importance is determined by interdependence, risk sensitivity, and spatial conflict. These three dimensions consider the network topology, node robustness, and transportation environment stability. A case study of the Belt and Road Initiative shipping network verified the validity and feasibility of this framework. The results indicated that the importance of nodes can be represented by their heterogeneity. Critical regions strongly coincide with the distribution of major global straits and transportation routes. Notably, the similarity of results under multi-features improves the accuracy of identifying critical nodes and regions within the complex network, whereas the differences compensate for the shortcomings of the single-dimensional approach. This provides actionable insights and guidance for stakeholders to build stability in maritime supply chains.