In this paper we will investigate a statistical modeling to perform the clustering of passengers based on their ticketing logs in the public transport. The aim is partition the passengers into groups on the basis of their travel hours. The clustering is proposed to be performed in unsupervised way by using advanced data partitioning tools, that is dedicated Gaussian mixture models. Doing so, we will be able to extract typical patterns describing different types of transport usage, namely sporadic usage, typical home-work commute behavior, scholar usage etc.