What do users want from transit? Qualitative analysis of current and potential users' perceptions

The study aims to unveil the common themes of public transit customers' opinions regarding improving transit service quality and, hence, develop an improvement framework for service quality grounded on customers' insights. The study utilized a dataset of 2,917 comments from current and potential public transit customers. The study employed a mixed-method approach that integrates text-mining-based topic modelling and grounded theory to investigate customers' comments and derive a customer-focused improvement framework. The Latent Dirichlet Algorithm (LDA) identified 42 latent themes in customers' comments and concluded that customers' concerns are highly associated with their familiarity with the service. The grounded theory approach concluded that customers' comments could be represented by 53 themes, 13 sub-categories and four major categories for current transit users. In comparison, 36 themes, eight sub-categories and three major categories are extracted for potential customers. The categories, sub-categories, and themes could be interpreted as an itinerary of actions that current and potential customers would like to see in order to improve their experience with the transit service. These findings offer an opportunity to guide quality improvement in a structured manner. It is worth mentioning that each of the adopted approaches (topic modelling and grounded theory) has its strengths and should be considered complementary.