Owing to the limited-time relevance and labour cost of traditional traffic service surveys, the quantity and quality of the data collected are not satisfactory in many cases. As more people are becoming active on comment websites and given the advances in natural language processing (NLP) techniques, it is feasible to conduct the opinion mining of comments regarding transportation services at large scale and low cost. In this study, >25,900 comments on websites regarding various public transportation services in Shanghai were analysed using NLP techniques, including sentiment analysis to calculate people's satisfaction with different transit facilities, keyword extraction to understand the key point in comments, and co-occurrence analysis to understand the commonality and characteristics of different routes. The sentiment analysis shows that people are more satisfied with traffic hubs than vehicles, and people have the highest level of satisfaction with airports and the lowest level of satisfaction with buses. The keywords classified by sentiment values show the strengths and weaknesses of Bus No. 49. Finally, the authors interpret a few anomalies of the results by tracing the context around the keywords. This study designs an NLP approach of transportation survey to improve the quality of public transportation service.


    Access

    Access via TIB


    Export, share and cite




    Appraisal System of Public Transportation Service Quality

    Yin, L. / Li, Q. / Hu, Y. et al. | British Library Conference Proceedings | 2009


    Passenger Satisfaction Scale for Public Transportation

    Kökalan, Özgür / Tutan, Adem | Transportation Research Record | 2020


    Appraisal System of Public Transportation Service Quality

    Yin, Liang / Li, Qingsong / Hu, Yanli | ASCE | 2009


    Predicting passenger satisfaction in public transportation using machine learning models

    Ruiz, Elkin / Yushimito, Wilfredo F. / Aburto, Luis et al. | Elsevier | 2024