Auteurs : Lilian Loubet ¹; Géraldine del Mondo ²; Julius Bañgate ²; Eric Sanlaville ²; Pierrick Tranouez ²

• 1 Identité et Différenciation de l’Espace, de l’Environnement et des Sociétés
• 2 Laboratoire d'Informatique, du Traitement de l'Information et des Systèmes

The paper presents a model designed to analyse port governance. It considers that thestakeholders ability to adopt cooperative behaviors constitutes the key element to portdevelopment. Its focus is on medium-sized European ports and fifteen cases were studied. Eachterritory was first subject to a qualitative survey and analysis of the contents of local stakeholdersdiscourse (over 80 interviews conducted). The material is rich, allowing for the comparisonbetween two or even three ports, yet the delicate nature of the relations brought to light addsconsiderable complexity to the comparison within a larger ensemble. The paper, therefore,proposes a semi-automatic treatment which helps to mitigate this difficulty by means of acomputer model based on graph theory. It involves a modelling system based on the relationsbetween the entities of the system. In this context, the relations between stakeholders wereanalysed in order to create typologies and eventually envisage some standard models ofgovernance. In order to territorialize the subject, six typical cases out of fifteen were used: LeHavre, Nantes-Saint Nazaire, Dunkirk (France), Gdynia (Poland), Klaipėda (Lituania), Hamina-Kotka (Finland).All these port environments modelled according to a single format (i.e. a graph) led to theapplication of a certain number of metrics which enables them to be compared. Two mainmetrics were presented in the framework of this paper for illustrative purposes: “Density” and“S_metric”. These metrics originating in graph theory, coupled with other indicators (distributionof degrees and number of hubs per port), allowed to measure the relationships' intensity and thedistribution of these intensities among the stakeholders, and to identify the main stakeholders orconversely the least influential.