Discovering Mobility Patterns on Bicycle-Based Public Transportation System by Using Probabilistic Topic Models

In this work, we present a new framework to discover the daily mobility routines which are contained in a real-life dataset collected from a bike-sharing system. Our goal is the discovery and analysis of mobility patterns which characterize the behavior of the stations of a bike-sharing system based on the number of available bikes along a day. An unsupervised methodology based on probabilistic topic models has been used to achieve these goals. Topic models are probabilistic generative models for documents that identify the latent structure that underlies a set of words. In particular, Latent Dirichlet allocation (LDA) has been used to discover mobility patterns. Our database has been collected for almost half a year from the Bicicas bike sharing system in Castellón (Spain). A set of experiments have been conducted to demonstrate the type of patterns that can be effectively discovered by using the proposed framework.