Exploring Travel Behavior and Activity Patterns using Urban Transit Mobility Sensing Data
Date
2023-12-13
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
In this study, we employ a probabilistic topic modeling algorithm, known as Latent Dirichlet Allocation (LDA), to autonomously deduce the purposes of trips based on activity characteristics extracted from smart card transit data. While the majority of existing literature has primarily concentrated on identifying patterns related to home and work-related activities, our research delves deeper into the realm of non-home and non-work activities, aiming to uncover distinctive patterns associated with a more granular spectrum of activities. Temporal attributes of activities are derived from trip data recorded by the Tehran subway's automatic fare collection system. Furthermore, we enrich the spatial attributes of non-home and non-work activities by incorporating land-use data. Multiple activity attributes, including start time, duration, frequency, and land-use information, are harnessed to infer activity purposes and patterns. Our analysis uncovers 14 distinct patterns associated with non-commuting activities, based on their temporal and spatial characteristics. These patterns encompass educational, recreational, commercial, health, and other service-related activity types. To gain further insights, we analyze changes in passenger trip patterns and behaviors before and during the COVID-19 pandemic, with a specific focus on non-home and non-work-related activities. Our investigation reveals significant alterations in these patterns. For instance, we observe a reduction in both the number and duration of recreational patterns, alongside the elimination of morning patterns in educational activities. Moreover, the number of commercial activities has decreased. The proposed model effectively captures shifts in travel behavior triggered by various disruptions, making use of smart card transit data. This capacity holds the potential to facilitate travel demand modeling, inform future planning and system management, and enable more adaptive decision-making processes.
Description
Keywords
Activity pattern, Topic modeling, Transit smart card
Citation
Aminpour, N. (2023). Exploring travel behavior and activity patterns using urban transit mobility sensing data (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.