A fuzzy logic model for taxi service assessment using passenger survey data from Ukraine and Poland

Taxi service quality is traditionally assessed using objective metrics like waiting time, cost, and comfort. However, contemporary research confirms that subjective psychological factors significantly influence service perception. This study develops an interpretable Mamdani fuzzy inference system (TEC model) that maps three passenger psycho-emotional factors—Trust, Empathy, and Control—into an interaction level (0–100) and a corresponding 1–5 star rating for taxi trips. The model is parameterized using a questionnaire survey in Ukraine (n = 118) and Poland (n = 101) and examined through a pilot case study of 10 real trips in Kharkiv and Gdansk. In the pilot, the model-based star ratings were largely consistent with app ratings; the relative error was 6.67% and is reported as a descriptive indicator only. The framework is intended for offline diagnostic assessment and scenario-based improvement planning rather than real-time prediction. Further validation on larger and more diverse samples is required to quantify generalizability. The novelty of this research lies in translating a passenger’s subjective psycho-emotional perceptions into quantitative indicators, which increases the reliability of taxi service quality assessment. The results are significant for taxi services and the logistics of their operations, providing a deeper understanding of the psychological factors that influence ratings. Furthermore, this approach can be used by digital taxi platforms for more effective service management. The work demonstrates the potential of fuzzy logic to improve predictive accuracy, develop personalized strategies, and create a foundation for intelligent decision support systems to enhance taxi services in European cities.