An analysis of tripadvisor reviews of 127 urban rail transit networks worldwide. (January 2022)

• Record Type: Journal Article
• Authors: Taecharungroj, Viriya
• Is Part Of: Travel behaviour and society. Volume 26(2022)
• Page Start: 193
• Page End: 205
• Publication Date: 2022-01
• Subjects: Urban rail transit -- Traveller experience -- Metro -- Subway -- LDA -- Textual analysis; Transportation -- Periodicals; Population geography -- Periodicals; 303.48305
• DOI: 10.1016/j.tbs.2021.10.007
• Languages: English
• ISSNs: 2214-367X

Abstract

Highlights: This research analysed 185, 485 reviews on TripAdvisor collected from 127 urban rail transit networks in 123 cities. A topic modelling algorithm, latent Dirichlet allocation (LDA), produced eleven experiential dimensions. Salience-valence analyses were performed to identify factors that influence positive and negative experiences. Performance, card, and communication are the most positive experiences; crowd and safety are the least positive. This research presents a rapid broad diagnostic tool that can be used by URTNs to manage, develop, and benchmark their services. Abstract: Urban rail transit networks (URTNs) are essential for sustainability, equity and liveability of cities worldwide. Recent studies have explored ways in which URTNs can improve traveller experiences; however, the lack of a comprehensive method to benchmark URTNs worldwide and the lack of travellers' voices and attitudes in conceptualisation call for a new technique to identify and assess URT traveller experience. This research aimed to uncover the experiential dimensions of URT travellers from TripAdvisor reviews and to compare URTNs worldwide. The research scope included 185, 485 reviews on TripAdvisor collected from 127 URTNs in 123 cities. Data were analysed using an unsupervised machine learning technique, latent Dirichlet allocation (LDA). The analysis found eleven experiential dimensions: card, ticketing, direction, communication, architecture, network, performance, safety, crowd, connectivity and traffic . Subsequent salience-valence analyses identified factors that drive positive and negative experiences. This study presents a rapid and insightful method to assess traveller experiences. URTNs worldwide can use this technique to analyse and improve their services.