The Digital Revolution has profoundly changed the tourism segmentation research field. It is now easier to grasp tourists' behaviors thanks to their digital traces left on social networks. The studies that have centered their method around tourists' digital traces focus on applying existing clustering algorithms to the tourism context. In this paper, we propose a measure to determine tourism segmentation — also known as tourism profiling — by establishing a new clustering algorithm focusing on stays conducted by tourists. This measure is based on both the context and the content of the trips. The approach is simulated and evaluated experimentally on a real dataset at various periods and on diverse nationalities, in the French capital Paris.
The data used in the Results section of our paper are not included in this repository due to their size. However, a sample of each dataset used is proposed in the data folder.