Page 60 - JCAU-6-3

Journal of Chinese
Architecture and Urbanism Age-friendly smart communities in Beijing

(ii) How can we identify and evaluate the level and distribution of the aging population and communities in
characteristics of a community’s age-friendliness? Beijing.
We anticipated that data analytics would shed light on To further explore the differences among aging
the location of existing communities, the spatial proximity communities and propose more suitable transformation
of dwelling units and amenities critical for quality of life, suggestions, we selected 12 key indicators covering various
and local demographic and socioeconomic conditions. aspects of community characteristics: Housing prices,
urban centrality, and house age, proportion of the elderly
3.2. Data collection population, commercial availability, cultural richness,
We gathered diverse urban data from various sources, food availability, health-care accessibility, leisure activities,
encompassing public data from the city’s open-data portal, public service availability, general service provision, and
application programming interface (API) request-derived transit convenience (Table 1). These indicators were used
data, cell phone signaling data, and website-acquired data. for classifying the characteristics of aging communities
Specifically, we extracted the geo-location (longitude through K-means cluster analysis, a widely accepted
and latitude) of 76,462 anonymous individuals through method for dividing data points by minimizing intra-cluster
cell phone signaling data. This dataset also provides age distance (Kamińska et al., 2023). The elbow method was
information, enabling us to discern the spatial distribution employed to identify the optimal number of clusters (i.e., K
of the elderly population in Beijing. Furthermore, we values). This method involves clustering different K values
collected real estate data from a third-party data platform, and evaluating the sum of squares of the total in-cluster
which reports critical details concerning residential errors corresponding to each K value. We identified the
neighborhoods, such as geographical coordinates, building point at which the increase in the sum of squares of error
age, street addresses, and property-related information. significantly decelerates, indicating the optimal K value.
In addition, we acquired datasets from the Beijing Open The application of this method aids in effectively identifying
Data Platform, including the spatial distribution of public natural groups in multidimensional data space, ensuring
service facilities and points of interest (POIs) collected the reliability and accuracy of clustering results. The goal is
from Gaode Maps API. to select the most suitable communities for transformation,
closely aligning with the living environment of the elderly,
3.3. Identification and characterization
and to use measurable methods to improve the quality
This study established the basic criteria for age-friendly of urban development in Beijing while also adapting
community rehabilitation based on the specific guidelines this analytical method for application in different urban
for community regeneration set forth by the Chinese environments (Mehaffy et al., 2020). Such areas are often
government and international standards for aging regarded as the most challenging urban environments due
communities. These criteria incorporate two major to the limited capacity of their residents to access external
considerations gleaned from previous studies: First, the services (Buffel & Phillipson, 2023).
community should have been established no less than
20 years ago (General Office of the State Council, 2020; 4. Results
van Hoof et al., 2022); and second, the proportion of 4.1. Spatial distribution of elderly population and
residents aged 65 and older in the community should be housing
7% or more (United Nations, 1956). These criteria consider
both the durability of the community’s infrastructure and Figure 2 summarizes the spatial distribution of the
the impact of an aging population on the evolving needs of elderly population and housing properties in Beijing.
the community functions and services (Kołat et al., 2022). When examining local characteristics in Beijing, the
Quantitative metrics were extracted from POIs within average housing price often serves as an indicator of
a 500 m-radius circular buffer to represent community neighborhood quality, considering various conditions
characteristics in accordance with WHO’s eight indicators and resources, such as location, transportation, school
of age-friendliness (World Health Organization, 2007) district, services, and facilities (Figure 2C). A comparison
and empirical cases (Garau et al., 2016; Sun et al., 2016). of the distribution of the elderly population and the
The POI data include health facilities, governmental and average housing price reveals no direct correlation
organizational services, public health amenities, cultural between these two factors. Noticeably, the southern part
facilities, commercial establishments, leisure venues, and of the city exhibits relatively low housing prices despite
transit options, among others. Table 1 summarizes each a higher concentration of elderly residents. We infer that
critical factor’s definition, data sources, and quantification while the total distribution of the elderly population
description. These indicators enable us to scrutinize the correlates strongly with the city’s development timeline,

Volume 6 Issue 3 (2024) 5 https://doi.org/10.36922/jcau.1754