@article{ART002139555},
author={Lee, Gunhak and Kamyoung Kim},
title={Estimating de facto Population Using Spatial Statistics},
journal={Journal of the Korean Cartographic Association},
issn={1598-6160},
year={2016},
volume={16},
number={2},
pages={71-93}
TY - JOUR
AU - Lee, Gunhak
AU - Kamyoung Kim
TI - Estimating de facto Population Using Spatial Statistics
JO - Journal of the Korean Cartographic Association
PY - 2016
VL - 16
IS - 2
PB - The Korean Cartographic Association
SP - 71
EP - 93
SN - 1598-6160
AB - Recently, demand for statistics of the potential service population is increasing for the analysis of trade area for small business establishment and the investigation of regional industrial demands. The census is not appropriate for estimating a timely potential service population because the census survey is usually implemented at a regular long interval. Mobile telecommunications big data collected from individual’s mobile phone enables to overcome the shortcomings of the census and presents the real-time dynamics of regional population distribution and flow. This research attempts to estimate a de facto population as potential service demand of the region using a mobile telecommunications big data and spatial statistics. For doing this, we first examined the distribution of the ‘service population’ of Daegu metropolitan city in terms of gender, age, and temporal periods, which is provided by a commercial mobile telecom company. Then, we estimated a de facto population by census output area using OLS and spatial regression models based on a number of explanatory variables regarding population, social, and economic characteristics of the region. The empirical results show that a potential service population can be significantly estimated using regional statistical data and the spatial regression model with an explicit consideration of spatial dependency of population is the best fit for de facto population estimation. The estimation error is smallest in the level of the census output area and de facto population by age group is more accurately estimated than other population categories. For the spatial distribution of the estimation error, over-estimation has often occurred in central areas of the city and under-estimation has appeared in outer areas of the city. This might be led by the regional differentiation of mobile-phone usage and mobility.
KW - De facto population;Service population;Mobile telecommunications big data;Spatial statistics;Spatial regression models
DO -
UR -
ER -
Lee, Gunhak and Kamyoung Kim. (2016). Estimating de facto Population Using Spatial Statistics. Journal of the Korean Cartographic Association, 16(2), 71-93.
Lee, Gunhak and Kamyoung Kim. 2016, "Estimating de facto Population Using Spatial Statistics", Journal of the Korean Cartographic Association, vol.16, no.2 pp.71-93.
Lee, Gunhak, Kamyoung Kim "Estimating de facto Population Using Spatial Statistics" Journal of the Korean Cartographic Association 16.2 pp.71-93 (2016) : 71.
Lee, Gunhak, Kamyoung Kim. Estimating de facto Population Using Spatial Statistics. 2016; 16(2), 71-93.
Lee, Gunhak and Kamyoung Kim. "Estimating de facto Population Using Spatial Statistics" Journal of the Korean Cartographic Association 16, no.2 (2016) : 71-93.
Lee, Gunhak; Kamyoung Kim. Estimating de facto Population Using Spatial Statistics. Journal of the Korean Cartographic Association, 16(2), 71-93.
Lee, Gunhak; Kamyoung Kim. Estimating de facto Population Using Spatial Statistics. Journal of the Korean Cartographic Association. 2016; 16(2) 71-93.
Lee, Gunhak, Kamyoung Kim. Estimating de facto Population Using Spatial Statistics. 2016; 16(2), 71-93.
Lee, Gunhak and Kamyoung Kim. "Estimating de facto Population Using Spatial Statistics" Journal of the Korean Cartographic Association 16, no.2 (2016) : 71-93.
