Infering the passenger’s trip purpose in the Smart card data using data mining techniques, an case study of the Belo Horizonte Brazilian city

Resumo

Planning a quality public transport system starts with collecting data about passenger demand. Tradi-
tional data collection forms are expensive and do not precisely represent the travel demand. On the other hand,

secondary data collected passively, for long and continuous periods, and at low cost is emerging as a new oppor-
tunity, such as Smart-cards data. However, these data are also limited and miss important information, such as

trip purpose. Knowing the trip purpose of public transport passengers is essential to ensure integrated planning
between transport and land use and to guarantee higher quality of the public transport service and attract more
users, thus contributing to the mitigation of excessive use of the car and its impacts. In this context, the process
of Knowledge Discovered in Databases can be used to extract knowledge from these secondary data, improving
their applicability in travel demand models. Under this perspective, this study contributes to the enrichment of
Smart cards data from the public transportation system of the Metropolitan Region of Belo Horizonte by infering
the passengers’ trip purposes using data mining techniques.