Increasing the Likelihood of Finding Public Transport Riders that Face Problems Through a Data-Driven approach

Abstract

The maintenance of big cities public transport service quality requires constant monitoring, which may become an expensive and time-consuming practice. The perception of quality, from the users point of view is an important aspect of quality monitoring. In this sense, we proposed a methodology for data analysis and visualization, supported by software, which allows for the structuring of estimates and assumptions of where and who seems to be having unsatisfactory experiences while making use of the public transportation in populous metropolitan areas. Moreover, it provides support in setting up a plan for on-site quality surveys. The proposed methodology increases the likelihood that, with the on-site visits, the interviewer finds users who suffer inconveniences, which influence their behavior. Simulation comparison and a small-scale pilot survey stand for the validity of the proposed method.