2021
Chen, Chao; Feng, Tao; Ding, Chuan; Yu, Bin; Yao, Baozhen
Examining the spatial-temporal relationship between urban built environment and taxi ridership: Results of a semi-parametric GWPR model Journal Article
In: JOURNAL OF TRANSPORT GEOGRAPHY, vol. 96, 2021, ISSN: 0966-6923.
Abstract | Links | BibTeX | Tags: Big data, Geography
@article{WOS:000703846700010,
title = {Examining the spatial-temporal relationship between urban built environment and taxi ridership: Results of a semi-parametric GWPR model},
author = {Chao Chen and Tao Feng and Chuan Ding and Bin Yu and Baozhen Yao},
url = {https://www.sciencedirect.com/science/article/pii/S0966692321002258},
doi = {10.1016/j.jtrangeo.2021.103172},
issn = {0966-6923},
year = {2021},
date = {2021-10-01},
urldate = {2021-10-01},
journal = {JOURNAL OF TRANSPORT GEOGRAPHY},
volume = {96},
abstract = {With the advance of intelligent transportation systems (ITSs) and data
acquisition systems (DASs), it becomes possible in recent to explore the
determinants of urban taxi ridership using multi-source heterogeneous
data. This paper aims to use floating car data, points-of-interests
(POIs) data and housing-price data to assess the influence of the built
environment on taxi ridership. Within a scale of 0.5 km grid, critical
indicators related to the economic aspect, intermodal connection, and
land use factors were obtained using the multi-source data in Shanghai.
To capture the spatial and temporal heterogeneity, Semi-parametric
Geographically Weighted Poisson Regression (SGWPR) models are built over
different time dimensions. It is found that SGWPR models result in
higher goodness-of-fit than the generalized linear models. More
importantly, the results show the impacts of built environment factors
on taxi demand are highly heterogeneous, positive or negative in
different city areas, reflected in the significant temporal variations
of the effects. Overall, these findings suggest that the built
environment factors have significant impacts on urban taxi demand, and
the spatial context should not be ignored. Findings in this paper are
expected to help better understand the relationship between urban taxi
demand and built environment factors, improving the service level of the
urban taxi system, and offering valuable insights into future urban and
transportation planning.},
keywords = {Big data, Geography},
pubstate = {published},
tppubtype = {article}
}
With the advance of intelligent transportation systems (ITSs) and data
acquisition systems (DASs), it becomes possible in recent to explore the
determinants of urban taxi ridership using multi-source heterogeneous
data. This paper aims to use floating car data, points-of-interests
(POIs) data and housing-price data to assess the influence of the built
environment on taxi ridership. Within a scale of 0.5 km grid, critical
indicators related to the economic aspect, intermodal connection, and
land use factors were obtained using the multi-source data in Shanghai.
To capture the spatial and temporal heterogeneity, Semi-parametric
Geographically Weighted Poisson Regression (SGWPR) models are built over
different time dimensions. It is found that SGWPR models result in
higher goodness-of-fit than the generalized linear models. More
importantly, the results show the impacts of built environment factors
on taxi demand are highly heterogeneous, positive or negative in
different city areas, reflected in the significant temporal variations
of the effects. Overall, these findings suggest that the built
environment factors have significant impacts on urban taxi demand, and
the spatial context should not be ignored. Findings in this paper are
expected to help better understand the relationship between urban taxi
demand and built environment factors, improving the service level of the
urban taxi system, and offering valuable insights into future urban and
transportation planning.
acquisition systems (DASs), it becomes possible in recent to explore the
determinants of urban taxi ridership using multi-source heterogeneous
data. This paper aims to use floating car data, points-of-interests
(POIs) data and housing-price data to assess the influence of the built
environment on taxi ridership. Within a scale of 0.5 km grid, critical
indicators related to the economic aspect, intermodal connection, and
land use factors were obtained using the multi-source data in Shanghai.
To capture the spatial and temporal heterogeneity, Semi-parametric
Geographically Weighted Poisson Regression (SGWPR) models are built over
different time dimensions. It is found that SGWPR models result in
higher goodness-of-fit than the generalized linear models. More
importantly, the results show the impacts of built environment factors
on taxi demand are highly heterogeneous, positive or negative in
different city areas, reflected in the significant temporal variations
of the effects. Overall, these findings suggest that the built
environment factors have significant impacts on urban taxi demand, and
the spatial context should not be ignored. Findings in this paper are
expected to help better understand the relationship between urban taxi
demand and built environment factors, improving the service level of the
urban taxi system, and offering valuable insights into future urban and
transportation planning.