Research Information System
International Journal of Transportation Science and Technology, 2025 (ESCI, Scopus)
This study investigates the factors influencing passenger satisfaction on the Tyne and Wear Metro system, a key public transportation network in North East England. The research explores how various service attributes, such as accessibility, reliability, and comfort, contribute to overall user satisfaction. Data from 850 passengers were collected through a structured questionnaire and analysed using advanced machine learning (ML) and deep learning (DL) techniques, including Random Forest, Gradient Boosting, and Neural Networks. This study addresses a gap in the literature by applying a multi-model ML/DL approach to a hybrid light rail system, incorporating sustainability-oriented variables such as environmental concern into the analysis of passenger satisfaction. The Random Forest model demonstrated high predictive accuracy, achieving Area Under the Curve scores of 0.93 and 0.91 for specific user classes. The findings highlight critical areas for improvement, particularly in service reliability and comfort, while also underscoring the potential of the Metro to increase ridership and reduce emissions through enhanced service quality. The findings can guide operators such as Nexus in prioritising service improvements, particularly in reliability and comfort, to boost ridership and support the region’s Net Zero targets. This research provides valuable insights for transit authorities aiming to enhance user satisfaction and promote sustainable urban mobility.