Discovering the hidden community structure of public transportation networks

TL;DR

This paper introduces two novel network structures derived from public transit contact data to uncover passenger community patterns and assess epidemic spreading risks within a metropolitan transit system.

Contribution

It develops transfer and community networks from transit contact data, providing new insights into passenger groups and travel behavior for epidemic risk modeling.

Findings

Transfer network reveals passenger groups traveling together.

Community network captures similarity in travel patterns.

Routes connecting to city center are high-risk during outbreaks.

Abstract

Advances in public transit modeling and smart card technologies can reveal detailed contact patterns of passengers. A natural way to represent such contact patterns is in the form of networks. In this paper we utilize known contact patterns from a public transit assignment model in a major metropolitan city, and propose the development of two novel network structures, each of which elucidate certain aspects of passenger travel behavior. We first propose the development of a transfer network, which can reveal passenger groups that travel together on a given day. Second, we propose the development of a community network, which is derived from the transfer network, and captures the similarity of travel patterns among passengers. We then explore the application of each of these network structures to identify the most frequently used travel paths, i.e., routes and transfers, in the public transit system, and model epidemic spreading risk among passengers of a public transit network, respectively. In the latter our conclusions reinforce previous observations, that routes crossing or connecting to the city center in the morning and afternoon peak hours are the most "dangerous" during an outbreak.