An Insight into the Kahramanmaraş Public Transportation System Using Statistical and Artificial Intelligence Methods

Creative Commons License

Kahveci M. Y., Sikar R. B., Aslan H.

International Cities4Youth Symposium, Sakarya, Türkiye, 11 Temmuz 2025, ss.1-3, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: Sakarya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.1-3
  • İstanbul Üniversitesi-Cerrahpaşa Adresli: Evet

Özet

This study investigates the transition of Kahramanmaraş's public transportation system from a fragmented minibus based model to a centralized municipal bus network, emphasizing its impact on passenger satisfaction. A structured 13-item Likert-scale survey was conducted with 85 participants at key transit hubs, and the results were analyzed using both SPSS and Artificial Neural Networks (ANN). Statistical analysis revealed improvements in service frequency and comfort, while concerns remained regarding access to travel information and complaint handling. The ANN model was developed to assess the relative importance of socio-demographic and behavioral variables in shaping satisfaction. Permutation-based feature importance analysis indicated that gender, particularly female respondents, was the most influential factor, followed by private vehicle ownership and educational background. These findings reveal pronounced differences in service perception across user groups. Notably, young passengers, especially university students, played a crucial role in shaping feedback due to their high share among public transport users. The study demonstrates the value of AI-based tools for identifying complex patterns in survey data and informing user-centered transit planning. By integrating statistical and machine learning approaches, this research offers practical insights for designing more inclusive and sustainable urban mobility strategies, not only in Kahramanmaraş but also in similar mid-sized metropolitan areas.