A model based on multiple one-period possibilistic Markov chains to simulate the tourist flow generated by a cruise ship docked in Palma's port

Abstract

[eng] This article studies how to simulate the tourist flow in a city from the point of view of the task allocation problem. We assume that all tourists start their tour from a cruise ship docked in the Port of Palma. Then, each tourist decides which places to visit and the order to do it. The tourist making decision process to choose the path depends on a combination of two stimulus, the utility to visit each place and the distance to go there. The utility is a subjective perception modeled by the stars and reviews downloaded from Google “Things to do” in Palma and the distance is computed by the Vincentry distance. Using several aggregation functions we merge this information to define the tourist response and, thus, to model how the tourist feels stimulated to visit the different places of the city over the time. The tourist decision making process is modeled by two different decision making methods, one of them solves a subjective optimization problem and the other one is based on sampling of possibility distributions. In order to generate the simulations, all the aforementioned ingredients are put in common by means of multiple one-period possibilistic Markov chains which provide the route chosen by tourists over time. The simulations and the model are implemented in Python.