Characterising and Predicting Urban Mobility Dynamics by Mining Bike Sharing System Data

Abstract

Understanding the spatio-temporal characteristics of human mobility in urban areas is invaluable especially for traffic management and urban planning. An opportunity to characterize and predict urban mobility is provided by mining Bike Sharing System (BSS) trip data spatially and temporally. This study focuses on identifying highly predictable BSS users, revealing their mobility characteristics and predicting their next-place movements. In undertaking user classification, we compare between naïve subscription based classification, namely registered and unregistered users, and trip number based classification, namely commuter, regular and casual users. Our analysis of 13 weeks of London BSS data shows that users with high predictability are those with 50 or more historical trips and with commuting patterns that can be fairly represented by registered users. However, because of the diversity of usage patterns, the spatio-temporal mobility characteristics of registered users are not as uniform as commuters who are the fastest, farthest travelled, most consistent, and most predictable riders with Markovian qualities. Using only a first order Markov predictor, the prediction of trip destination based on trip origin in weekday morning is up to 78% predictable for commuters.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.