Chapter 7 - Uncovering spatiotemporal structures from transit smart card data for individual mobility modeling

Abstract

While conventional travel survey data are limited in sample size and observation period, recent advances in urban mobility sensing technologies afford the opportunity to collect traces of individual mobility at a large scale and over extended periods of time. As a result, individual mobility has become an emerging field dedicated to extracting patterns that describe individual movements in time and space. This chapter discusses the problem of individual mobility modeling, the characteristics of new mobility data, and how we may use such data to better model individual mobility. It also presents recent efforts in statistical approaches to extracting meaningful travel-activity patterns and behavioral insights from individual-level longitudinal travel records. Specifically, we focus on three problems related to the spatiotemporal structure in individual mobility—next trip prediction, latent activity inference, and pattern change detection. The proposed models and some preliminary results are summarized. Pseudonymized transit smart card data from London's passenger rail network are used as a case study for this analysis. The proposed approaches are general and can be adapted for other data sources.