Home // GEOProcessing 2015, The Seventh International Conference on Advanced Geographic Information Systems, Applications, and Services // View article

Mining Spatial and Temporal Movement Patterns of Passengers on Bus Networks

Authors:
Chunjie Zhou
Pengfei Dai
Fusheng Wang
Renpu Li

Keywords: movement pattern; traveling time; attractive areas; passenger demand; hierarchical clustering

Abstract:
The analysis of human behavior is the basis of understanding many social phenomena. Accurate and reliable human movement pattern mining can lead to instructive insight to transport management, urban planning and location-based services (LBS). As one of the most widely used forms of transportation, buses can tell a lot of stories about people, including passenger demands, areas people are interested in crossing each day, and their travel patterns. Based on a large database from a real bus system, this paper aims to mine spatial and temporal movement patterns of passengers: evaluating traveling time of passengers, predicting number of passengers to estimate passenger demand and the crowdedness in the bus, and identifying attractive areas for passengers. There are major challenges for mining human movement patterns on bus networks: inhomogeneous, seasonal bursty periods and periodicities. In this paper, we take a Poisson process approach to model and evaluate traveling time of passengers, which can reflect the time features of individuals and activity cycles among areas. To overcome the challenges, we propose three prediction models and further take a data stream ensemble framework to predict the number of passengers. To obtain meaningful patterns of attractive areas, we provide a hierarchical clustering based approach to group spatiotemporally similar pick-up and drop-off points, as people's interests to these areas vary significantly on time, days and seasons. Our performance study based on a real dataset of five months' bus data demonstrates that our approach is quite effective: among 86,411 passenger demands on bus services, more than 78% of them are accurately forecasted.

Pages: 94 to 99

Copyright: Copyright (c) IARIA, 2015

Publication date: February 22, 2015

Published in: conference

ISSN: 2308-393X

ISBN: 978-1-61208-383-4

Location: Lisbon, Portugal

Dates: from February 22, 2015 to February 27, 2015