A decision tree approach utilizing Quadrant analysis for satisfaction in high-speed rail systems

Abstract

Satisfaction is an essential parameter for a public transport service provider, mainly if providers compete. For this reason, decision-makers try to identify and improve the factors affecting passenger satisfaction, which is important not only for the profitability of the service but also for sustainable and environmentally friendly transportation. In this study, 803 passengers were asked to rate their expectations and perceptions about the High-Speed Rail Systems they experienced between Ankara and Eski,cehir in Turkey. By using the obtained data, a methodology was established to predict factors affecting satisfaction. The Classification and Regression Tree was used for the analysis, which does not require any predefined baseline relationship between dependent and independent variables. The quadrant analysis is utilized to determine a single input to represent expectations and perceptions instead of using them separately, which allows the comparison of expectations and perceptions. Also, the problem experiencing rate, the decisive component of the Impact Score technique, is included in the model as an input. As a result, passengers'satisfaction is estimated with an accuracy rate of 88.79%, and the factors affecting satisfaction were determined as Service Delivery, Fare Level and Type, Passenger Information, Accessibility, and Security.