We have mainly focused on cooperative behaviors in two types of concrete networks: the first is the exchange network that museums form to circulate the collections of works of art, and the second a network of theaters that organize festivals in [8]. The econometrics of the networks by the relative 'quasi-spatial' arrangement of the data is problematic and we have tested a model with latent space that allows the identification of the parameters of a gravitational model. The museums in contact in this model are museums related to the NMR (Meeting of National Museums). The only variable considered is the number of artworks exchanged. The proximity in the latent space between the museums reflects the "gravitational force" expressed in the exchanges.
In the second study [1], we were able to produce a cooperative game class 'festival games' whose concept we applied to the data collected on festivals. We provided a theoretical model that takes into account the flow of spectators who are returned from one show to another, as part of the festival. This allows to interpret the data while answering a theoretical questioning (convexity of the game class, value of Shapley, implementation of the solution). The solution provides a pricing and revenue allocation rule that responds to Shapley's axioms, including the principle of equal treatment of equals. This rule helps to stabilize coalitions. The implementation allows to apply the rule concretely if we know the flows.
Contribution of the axis: Form and particular properties of certain networks of cooperation. Evolution of the "stellar" configuration of the network. This is a rare example of a concrete application of cooperative game theory to a dataset. Revenue sharing rule that stabilizes cooperation. Definition of a cooperative game class: festival games and adapted pricing.