Predicting the Intention of Public and Private Vehicle Users to Shift to Public Transport Using Data Mining Techniques

Abstract

With increasing urbanization and the rising burden of private vehicle usage, encouraging a modal shift toward public transportation is essential for sustainable urban development. This study investigated the factors influencing the intention of commuters to shift to public transport by applying data mining techniques to behavioral and demographic data. A structured questionnaire collected data on perceptions of safety, comfort, time efficiency, convenience, and socioeconomic variables in the City of Kolkata, India. 345 questionnaires were considered, which were collected from public and private vehicle users. Using the WEKA tool, the PART decision tree classifier was employed to extract interpretable rules that explain commuter intentions. The study found that positive perceptions, travel convenience, and safety are important factors of future public transport usage, while youths having low income and adults show unwillingness unless quality conditions are satisfied. The study offers practical insights for transport policymakers to design targeted strategies that enhance service quality and influence commuter attitudes. The findings demonstrate the effectiveness of rule-based machine learning in predicting travel behavior and support the formulation of data-driven mobility policies.