The aim of this paper is to detect port maritime communities sharing similar international trade patterns, by a modelisation of maritime traffic using a bipartite weighted network, providing decision-makers the tools to search for alliances or identify their competitors. Our bipartite weighted network considers two different types of nodes: one represents the ports, while the other represents the countries where there are major import/export activity from each port. The freight traffic among both types of nodes is modeled by weighting the volume of product transported. To illustrate the model, the Spanish case is considered, with the data segmented by each type of traffic for a fine tuning. A sort of link prediction is possible, finding for those communities with two or more ports, countries that are part of the same community but with which some ports do not have yet significant traffic. The evolution of the traffics is analyzed by comparing the communities in 2009 and 2019. The set of communities formed by the ports of the Spanish port system can be used to identify global similarities between them, comparing the membership of the different ports in communities for both periods and each type of traffic in particular.

中文翻译：

---

识别港口海事社区：在西班牙案例中的应用


本文的目的是通过使用双向加权网络的海上交通模型来检测具有相似国际贸易模式的港口海运社区，为决策者提供寻找联盟或识别竞争对手的工具。我们的二分加权网络考虑两种不同类型的节点：一种代表港口，另一种代表每个港口有主要进出口活动的国家。两种类型节点之间的货运流量是通过对运输的产品量进行加权来建模的。为了说明该模型，考虑了西班牙的情况，并按每种类型的流量对数据进行分段以进行微调。一种链接预测是可能的，为那些拥有两个或更多港口的社区找到属于同一社区但某些港口尚未有大量交通的国家。通过比较 2009 年和 2019 年的社区来分析流量的演变。西班牙港口系统的港口形成的社区集可用于识别它们之间的全球相似性，比较两个社区中不同港口的成员资格特别是时段和每种类型的交通。