Analysing travel satisfaction of tourists towards a metro system from unstructured data

Document Type

Journal Article

Publication Date


Subject Area

mode - subway/metro, place - europe, place - urban, planning - methods, planning - service improvement


Travel satisfaction, Trip Advisor, Sentiment analysis, Paris metro, Latent Dirichlet Allocation, Artificial neural networks


Tourism promotes economic development in cities, where the quality of public transport has a direct impact on the attractivity of a destination, making important to understand how to improve the service they provide.

This study proposes a methodology to analyse and predict travel satisfaction of tourists towards an urban metro system using online reviews and uses TripAdvisor data and the Paris metro as a case study. The Latent Dirichlet Allocation method is used to extract the dimensions of the online reviews and several machine learning models are trained to find the one that most accurately predicts the tourists' travel satisfaction.

According to the results for our case study, TripAdvisor online reviews of tourists towards the Paris metro can be classified in three dimensions: “General”, “Ticketing System” and “Utility and Accessibility”, and an artificial neural network can predict the satisfaction of tourists towards this urban metro system.

This methodology can be easily applied to different online platforms and urban metro systems and can be used by metro managements in order to deliver a better service to improve satisfaction among the tourists using their metro systems, and develop strategies aiming to improve the services they provide to attract more tourists to the cities.


Permission to publish the abstract has been given by Elsevier, copyright remains with them.


Research in Transportation Business & Management Home Page: