Data Mining to Build the OD Matrix in Developing Countries: Preliminary Analysis in an Argentinean Case Study

Abstract

The design of efficient public transportation systems is one of the main con-cerns for city administrators. These systems are usually very costly, and cities have to devote a significant portion of their public budgets to ensure the qual-ity of service. Thus, cost reduction is a relevant issue, in particular in the case of developing countries. On top of that, an increasingly relevant goal is the environmental sustainability of the system. In this contribution we address these two goals, studying them in the case of Bahía Blanca, a city in southern Argentina. We develop a procedure to gather data from the smart cards of us-ers of the system to compute a system's Origin-Destination matrix (OD).