sustainability

Article
Identifying Spatial Patterns of Retail Stores in Road
Network Structure
Zhigang Han 1,2,3,4 , Caihui Cui 2,4, *, Changhong Miao 5 , Haiying Wang 1,2,4 and Xiang Chen 6
1 Key Laboratory of Geospatial Technology for the Middle and Lower Yellow River Regions (Henan
University), Ministry of Education, Kaifeng 475004, China
2 College of Environment and Planning, Henan University, Kaifeng 475004, China
3 Institute of Henan Spatio-temporal Bigdata Industrial Technology, Henan University,
Zhengzhou 450046, China
4 Urban Bigdata Institute, Henan University, Kaifeng 475004, China
5 Key Research Institute of Yellow River Civilization and Sustainable Development, Henan University,
Kaifeng 475004, China
6 Department of Geography, University of Connecticut, Storrs, CT 06269, USA
* Correspondence: chcui@henu.edu.cn




Received: 26 June 2019; Accepted: 16 August 2019; Published: 21 August 2019


Abstract: Understanding the spatial patterns of retail stores in urban areas contributes to effective
urban planning and business administration. A variety of methods have been proposed in the
scientific literature to identify the spatial patterns of retail stores. These methods invariably employ
arbitrary grid cells or administrative units (e.g., census tracts) as the fundamental analysis units.
As most urban retail stores are distributed along street networks, using area-based analysis units is
subject to statistical biases and may obfuscate the spatial pattern to some extent. Using the street
segment as the analysis unit, this paper derives the spatial patterns of retail stores by crawling
points of interest (POI) data in Zhengzhou, a city in central China. Then, the paper performs the
network-based kernel density estimation (NKDE) and employs several network metrics, including
the global, local, and weighted closeness centrality. Additionally, the paper discusses the correlation
between the NKDE value and the closeness centrality across different store types. Further analysis
indicates that stores with a high correlation tend to be distributed in city centers and subnetwork
centers. The comparison between NKDE and cell-based KDE shows that our proposed method can
address potential statistical issues induced by the area-based unit analysis. Our finding can help
stakeholders better understand the spatial patterns and trends of small business expansion in urban
areas and provide strategies for sustainable planning and development.

Keywords: POI; road network; kernel density estimation; closeness centrality

1. Introduction
Location is key to the success of the brick-and-mortar retail business. “Location, location, location,”
as one of the oldest mantras in business planning and investment, has often been the primary business
consideration when selecting an operating site. An excellent location may be enough to make a retailer
successful even if its strategy is mediocre. Similarly, a poor location can be a significant burden that
even a good retailer cannot overcome [1]. Finding the optimal location for a retail store and conducting
retail location analyses have been longstanding aims of both retail business and urban planning. In the
process of urban growth, the expansion of the city is increasingly dependent on the city’s function as
a consumer hub [2,3]. The spatial configuration of the urban retail industry is of great significance
for maximizing the economic benefits and optimizing the transport system [4]. Understanding the

Sustainability 2019, 11, 4539; doi:10.3390/su11174539 www.mdpi.com/journal/sustainability

Sustainability 2019, 11, 4539 2 of 20

spatial distribution of commercial facilities is essential to the future planning of urban commercial retail
systems and contribute to the efficient design of urban commercial space. Aiming at the sustainable
development of the city, this improved understanding can avoid squandering planning resources that
could be better allocated to other needs.
Traditional retail location analysis focuses on the trading-area and site selection of individual
retail stores. These studies often use Reilly’s law or Huff’s model to infer the trading-area of stores
or facilities [5,6] to determine the location of a restaurant in conjunction with the location-allocation
model [7] in the context of Geographical Information Systems (GIS) [8]. Nevertheless, a single retail
store “does not tell the whole story” [9]. From the perspective of spatial planning, the location of
various types of retail stores determines the retail business structure of the city, which could be
leveraged to uncover the market potential. Many cases have examined the spatial structure of retail
commerce. Based on the central place theory, early studies defined the market area in a top-down
hierarchical system, spatially forming a business center network that is triangular in shape and contains
hexagonal market areas [10]. In addition, market, traffic, and public administration principles are used
to shape diversified distribution systems [11]. Thus, the commercial spatial structure becomes a more
diversified pattern when considering these non-spatial dimensions [12].
Given an increasing number of spatial analysis methodologies, most studies focused on the
location of retail stores by applying GIS techniques. For example, Porto et al. studied commercial and
service activities in Bologna, Italy. They discovered that retail and service activities in Bologna were
generally concentrated in areas that enjoyed better centrality [13]. Similar conclusions were reached
in a case study involving the correlation between retail stores and street networks in Barcelona [14],
Spain and in Changchun [15], Guangzhou [16], and Wuhan, China [17]. By utilizing a grid cell as the
analysis unit, the studies revealed the law of interaction between the location-distribution pattern of
urban retail stores and the structure of street networks. However, there are noticeable statistical biases
when using administrative units. First, the area-based unit analysis is subject to the classic modifiable
areal unit problem (MAUP), referring to the effect that both the analysis scale (i.e., the smallest unit
under observation) and the study scope considerably affect the statistical outcome. Second, using
area-based units is subject to the “edge effect” introduced by the relative placement of urban retailers:
In cities, the majority of stores are distributed along transport arterials, which are typically unit
boundaries. Thus, comparing to random distributions, the spatial pattern of stores summarized by
area-based units can be largely affected by the spatial organization of units [18]. To this end, it is better
to characterize the spatial pattern of retail stores using network-based structure.
Traditionally, streets are part of the public transportation network, and thus, people pay more
attention to their pass function. Since diverse urban economic activities occur on main streets,
they involve not only traffic but also trading, peddling, and selling in a more general sense, as when
Jacobs argued that streets are the lifeblood of cities rather than mere traffic channels [19–22]. As a
public space that carries traffic, economic activities, and other events, streets connect urban functions
physically and cognitively [22]. Early research on streets concentrated on road networks to explore
an extension of the Huff model [23] or the space syntax analysis [24,25]. Itzhak et al. [26] applied
the spatial syntax method to study the correlation between the spatial pattern of retail activities
and the spatial configuration of road networks in eight Israeli cities. With the progress of network
science, road network structures based on complex networks are receiving increasing levels of public
attention. Wang et al. [27] took population census blocks as the analysis unit to quantitatively analyze
the correlation between street centrality and land use intensity in Baton Rouge. In terms of land
use types, Wang et al. [28] explored the relationship between street centrality and different types of
urban land use in Shenzhen, China. Streets are also a key element in the urban image theory [29].
A series of cases have recently been conducted, including cases based on online street views [30–34].
By leveraging deep learning technology, Li et al. calculated the shadows of street trees in downtown
Boston by determining the Sky View Factor (SVF) from street images [35]. They then analyzed the
spatial distribution of solar radiation in the street canyons [36]. These studies reflect the unique value
Sustainability 2019, 11, 4539 3 of 20

of the street as a line cell in urban research. To date, there are very few such studies employing the
spatial structure of street networks for retail analysis.
In recent years, along with the rapid growth of mobile location-based service (LBS) technologies,
a large amount of multi-source LBS data can be obtained from the public domain and social
media. These data include points of interest (POIs), social network records, and cell phone signal.
In addition to reflecting the human dynamics in cities, LBS data can be used to reveal the first-order
distribution (e.g., economic indices, population intensity, condition of public facilities) and second-order
characteristics (human movements, flow of goods, social ties) of urban properties to better understand
the urban function, urban structure, and characteristics of crowd movement within a city [37,38].
For example, Wu et al. [39] studied and modeled crowd movement patterns of more than 15 million
check-in records in Shanghai. Pucci et al. [40] analyzed urban residents’ spatiotemporal patterns and
commuting characteristics in Milan, Italy, based on mobile signal data. There are rich semantic features
for POI data, which could uncover the activity space in a city. Coupled with machine learning methods
like topic modeling and word vector, employing POI data are able to identify urban functional areas
with added semantic information [41,42] and classify urban land use with the support of remote
sensing imagery [43,44].
Currently, POI data is widely used for urban studies, but the analysis unit is dominated by
area-based grid cells. Streets and their near vicinity are the primary space for commercial activities to
take place. Uncovering spatial patterns of retail stores in street unit can address potential statistical
issues induced by the area-based unit analysis. Thus, studies employing street networks play an
important role in the evaluation of the retail industry aimed at sustainable development [45,46]. In this
paper, we take Zhengzhou, a city in central China, as an example to analyze the distribution patterns
of different types of retail stores by using the street as the analysis unit. The paper also explores the
correlation between the distribution of retail stores and the street centrality indicator. In terms of the
spatial distribution pattern, we applied the network-based kernel density estimation to analyze the
spatial distribution of six types of retail stores. After creating the road network data set, we calculated
the global, local, and weight closeness centrality indicators separately. By focusing on a street as a linear
analysis unit, this paper aims to reveal more detailed spatial distribution rules of different types of retail
stores and to explore their interrelated characteristics with street closeness centrality. These multiple
tiers of analyses can shed light on urban commercial facility planning and sustainable development.
The paper is organized as follows. Section 2 details the methodology of network-based kernel
density estimation and the formation of closeness centrality. Section 3 describes our study area and data.
Section 4 introduces the statistical and spatial distribution of different types of retail stores in a road
network structure. Section 5 concludes the paper with implications for sustainable urban planning.

2. Methodology

2.1. Network-Based Kernel Density Estimation

The Kernel Density Estimation (KDE) method derives the density of the observations in terms of
the continuous probability distribution of the point-based data [47]. The KDE estimates the density
value at each location as the average value within a spatial window (i.e., the bandwidth) based on a
kernel function. The kernel function is a continuous normalized function that is centered at a particular
point and summarizes values within a given bandwidth [48], as defined by Equation (1).
n
1 X x − xi
 
ˆ( )
f x = 2 K (1)
h i=1 h

where h is the bandwidth and x-xi indicates the distance between the center of the kernel (x) and the
location xi . There are different types of kernels used for KDE. Here the popular quartic function is
employed [15], as given by Equation (2).
Sustainability 2019, 11, 4539 4 of 20

  2
3π−1 1 − vT v

 vT v < 1
k (x) =  (2)
0, otherwise


where v is di /h (di is the distance from the kernel center), and T indicates the transpose of the matrix.
Typically, the KDE uses grid cells as the analysis unit and the Euclidean distance as the distance
metric. As the urban space is not homogenous because of the mixed land use and the structure of road
networks, other forms of analysis units (e.g., census tracts) and distance metrics (e.g., the shortest path
distance) have been applied to real-world density assessments.
The network-based kernel density estimation (NKDE) method is the network extension of the
two-dimensional KDE. The distance metric employed by the NKED method is the shortest path
distance, which avoids the oversimplified Euclidean distance measure. In addition, the topological
relationship of the line elements in the network is considered. To avoid the overestimation of the
density and improve the computational efficiency, the equal-split kernel function method is introduced
to normalize the estimation results [49]. Suppose that the shortest path from x to xi contains p nodes:
v1 , . . . , vp and that ni represents the degree of the node vi , the form of the equal-split kernel function
for the network-based kernel density estimator [50] is defined by Equation (3).
 
x−x

 k h i /((n1 − 1)(n2 − 1) · · · · · · (ns − 1)), x − xi ≤ h


K (x) =  (3)

 0, x − xi > h

Stores may overlap in certain urban areas, such as near the central business district. To ensure
the accuracy of the measure, overlapped store locations were treated as the same point. The point
was weighted by the number of overlaps using the quantile mapping method, which could avoid
the skewed distribution of stores in highly clustered areas [51]. Then the kernel density value was
weighted for calculation. This method sorted the number of overlapped stores in ascending order, the
the ranks were used to compute the normalized weight value. Given the store set as c, the weight is
calculated by Equation (4).
(k − 1)(rmax − rmin )
wk = + rmin (4)
mc − me − 1
where wk is the weight of the kth ranked store location (from the smallest to the largest), rmax and rmin
are the user-defined parameters that control the influence of the weights on the kernel estimator, and
they are set to 2 and 1, respectively; mc and me are the total count of the c and the number of equal
values in c, respectively.

2.2. Closeness Centrality Index

The concept of closeness centrality has been used in the complex network theory [52,53]. It measures
how central a node is with respect to other nodes in the network [54]. Urban roads are a typical spatial
network which can be modeled by two methods. The first method is the primal network: The method
defines the road intersection as nodes; then, the road segment with the real length is abstracted into
the edge of the network. The other method is the dual network: The method abstracts the roads into
nodes and the connection between road segments into edges of the network. Comparing to the dual
network method, the primal network is able to retain the complete spatial network characteristics,
such as position and distance, and is invariably adopted by most GIS [55]. For this reason, this paper
employs the primal network and calculates the index of the closeness centrality (Cci ) to measure the
characteristics of the network nodes and edges. The closeness centrality is defined by Equation (5) [13].

N
X
Cci = (N − 1) / dij (5)
j=1, j,i
Sustainability 2019, 11, 4539 5 of 20

where N is the total number of nodes in the network, and dij are the shortest path distance between
nodes i and j.
The closeness centrality can be classified into the global closeness centrality (GCC) and the local
closeness centrality (LCC). The former calculates all the nodes in the network, while the latter only
involves nodes around a target node. To derive the LCC, an arbitrary distance value is usually applied
to estimate the buffer of the node. For example, in the case of Barcelona, Porta et al. [14] defined three
neighborhoods with
Sustainability 2019, 11, 4539a distance buffer of 800 m, 1600 m, and 2400 m, respectively. To overcome this
5 of 20
arbitrary nature, a method based on the modularity indicator is used to define the buffer, generating
overcome this arbitrary
network-based communities nature, a method based
(or subnetworks). Theon the modularity
modularity indicator
Mc quantifies is good
how used atocommunity
define the
buffer,
or generating
a network partitionnetwork-based
is, as given bycommunities
Equation (6) (or [56].subnetworks). The modularity 𝑀 quantifies
how good a community or a network partition is, as given by Equation (6) [56].
nc  !2 
 Lc kc 
X
Mc =  𝐿 − 𝑘  (6)
𝑀 = c=1 L − 2L (6)
𝐿 2𝐿

where n𝑛c isisthe

where thenumber
numberofof subnetworks,
subnetworks, Lc 𝐿is the number
is the numberof edges
of edgeswithin a subnetwork,
within a subnetwork, kc is 𝑘theistotal
the
degree of nodes in the subnetwork, and L is the total number of edges
total degree of nodes in the subnetwork, and 𝐿 is the total number of edges in the network.in the network. There are
There
many modularity-based
are many modularity-based community
community detection
detection algorithms
algorithms(e.g.,
(e.g.,Fast
Fastunfolding,
unfolding,Combo,
Combo, Informap,
Informap,
and
and Label propagation.) According to the study of Huang et al. [57], the Combo algorithm
Label propagation.) According to the study of Huang et al. [57], the Combo algorithm is is more
more
suitable for detecting spatial network communities. With the modularity
suitable for detecting spatial network communities. With the modularity value being the value being the optimization
objective, thisobjective,
optimization algorithmthis
discovers communities
algorithm through threethrough
discovers communities optimization strategies: Community
three optimization strategies:
division, community merging, and intercommunity node shift
Community division, community merging, and intercommunity node shift [58]. [58]. The weight of the
Theedge is defined
weight of the
by the number of bus lines between nodes. The road network in the study area
edge is defined by the number of bus lines between nodes. The road network in the study area is thusis thus divided into
30 subnetworks
divided with a module
into 30 subnetworks value
with of 0.9046
a module by the
value Combobyalgorithm
of 0.9046 the Combo (Figure 1). (Figure 1).
algorithm

Figure 1.
Figure 1. Result
Result of
of the
the community
community detection
detection in
in the
the study
study area.
area. The
The road
road network
network is
is divided
divided into
into 30
30
network-based communities by the Combo algorithm.
network-based communities by the Combo algorithm.
g
canbe
It can beseen
seenthat Ci 𝐶represents
that represents the global
the global feature
feature of the of the in
nodes nodes in the network,
the network, while 𝐶
while Cli describes
describes
the the local features.
local features. Because Because a roadboth
a road serves serves
theboth the (relative
global global (relative to the entire
to the entire city)the
city) and and the
local
g
local (relative to the subnetwork) functions, the weighted value of the 𝐶l and 𝐶
(relative to the subnetwork) functions, the weighted value of the Ci and Ci is used as an aggregate is used as an
aggregateofmeasure
measure of thecentrality
the closeness closenessof
centrality
the node,ofasthe node,inasEquation
shown shown in(7) Equation
[59]. (7) [59].
𝐶 = (1 − 𝑢 ) ∗ 𝐶 + 𝑢 ∗ 𝐶 , 𝑢 = 𝑒𝑐 ⁄𝑒𝑐 (7)
where 𝑢 , 𝑒𝑐 are the weight and the edge counts of the subnetwork m, respectively; 𝑒𝑐 is the
inter edge counts between the subnetwork m and other subnetworks. The weighted closeness
centrality of each node is calculated separately by Equation (7), and the average centrality value of
the two nodes of the road segment is generated as the closeness centrality of the road segment. Here,
Sustainability 2019, 11, 4539 6 of 20

g
Ci = (1 − um ) ∗ Cli + um ∗ Ci , um = ecinter
m /ecm (7)

where um , ecm are the weight and the edge counts of the subnetwork m, respectively; ecinter
m is the inter
edge counts between the subnetwork m and other subnetworks. The weighted closeness centrality of
each node is calculated separately by Equation (7), and the average centrality value of the two nodes of
the road segment is generated as the closeness centrality of the road segment. Here, the Urban Network
Analysis tool developed by Sevtsuk et al. [60] was used to calculate the node closeness centrality.

3. Study Area and Data

Henan province is the most populous province in China, and its capital, Zhengzhou, is the political,
economic, technical, and educational center of over 100 million people in central China. It is located on
the south bank of the middle reaches of the Yellow River (Figure 1). It is found in China’s hinterland
and is known colloquially as “the center of sky and earth.” The geographical coordinates of the city
range from 112◦ 420 –114◦ 140 E and 34◦ 160 –34◦ 580 N. The total area of Zhengzhou is 7446 km2 with an
urban area of 1010 km2 and a built-up area of 549.33 km2 . The economy of Zhengzhou ranks third in
central China and 17th among all Chinese cities in terms of GDP. In 2016, Zhengzhou was identified
as one of the fastest-growing cities in China [61]. Moreover, Zhengzhou is located at the intersection
of the national ordinary railway network and the high-speed railway network. The transportation
advantages have made Zhengzhou a commercial and trade mecca in central China.
Our study area is the contiguously urbanized built-up area of Zhengzhou. The area is situated
in central Zhengzhou, bounded on the north, east, west, and south of the 4th ring road of the city,
as shown in Figure 1. This area was selected because of the large concentration of retail stores in
the area. We extracted POI data using the application programming interface (API) from Baidu
Maps (map.baidu.com) and Gaode Maps (www.amap.com), which are two primary mapping service
providers in China, to construct a retail store dataset. Specifically, in order to obtain records for all
retail stores, we generated regular grid rectangles as the querying region. The rectangle is 15” in width
and 12” in height. There are 4133 rectangles in the study area. Each record collected contains a store’s
name, address, and location coordinates.
We then divided the retail stores into six broad categories consisting of specialty stores (SS),
department stores (DS), supermarkets (SMS), furniture stores (FS), construction material stores (CMS),
and consumer product stores (CPS), which mirrors the classification system used by the Chinese
government [12]. A specialty store is defined as a store that sells specific goods like clothes, footwear,
electronic products, etc. It is further refined into seven subtypes such as office supply stores (OSS),
apparel shop stores (AOS), home appliance stores (HAS), drug stores (DGS), car stores (CAS), cosmetic
stores (CTS), and other stores (OTS). The spatial datasets of the retail stores and street network were
built with ESRI ArcGIS. The final data set includes 78,777 retail stores (Table 1), 3523 network nodes,
and 5569 network edges of the street with the red line in the study area, as illustrated in Figure 2.

Table 1. Descriptive summary of retail stores in the study area.

Average Stores Line Cell Average Stores
Store Type Store Subtype Store Count Road Count
Per Street Count Per Line Cell
Office supply stores (OSS) 4285 897 4.7770 1374 3.1186
Apparel shop stores (AOS) 15456 1245 12.4145 1829 8.4505
Home appliance stores (HAS) 4919 1035 4.7527 1626 3.0252
Specialty
Drug stores (DGS) 3142 979 3.2094 1555 2.0206
stores
Car stores (CAS) 3811 434 8.7811 795 4.7937
(SS)
Cosmetics stores (CTS) 3569 860 4.1500 1367 2.6108
Other stores (OTS) 6587 1491 4.4178 2156 3.0552
Subtotal 41769 2285 18.2796 3335 12.5244
Department stores (DS) 249 121 2.0579 179 1.3911
Supermarkets (SMS) 2377 823 2.8882 1375 1.7287
Furniture stores (FS) 4211 492 8.5589 803 5.2441
Construction material stores (CMS) 13271 1147 11.5702 1887 7.0329
Consumer product stores (CPS) 16900 1966 8.5961 2967 5.6960
All Stores (AS) 78777 2652 29.7048 3934 20.0247
Sustainability 2019, 11, 4539 7 of 20
Sustainability 2019, 11, 4539 7 of 20

Figure 2.
Figure 2. Road
Road networks
networks of
of Zhengzhou,
Zhengzhou, China
China (2018).
(2018).

4. Results and Discussion

4.1. Descriptive
4.1. Descriptive Statistics
Statistics for
for Retail
Retail Stores
Stores
The results
The results including
includingthe thedescriptive
descriptivestatistics
statisticsofof stores,
stores, roads,
roads, andandlineline cells
cells (with
(with a length
a length of
of 400
400where
m m where the the stores
stores areare located)
located) areare given
given in in Table1.1.There
Table Thereareare41,769
41,769SS SSdistributed
distributed on on 2285
2285 roads
roads
and within 3335 line cells, with 18.28 stores on each road and 12.52 stores in each
and within 3335 line cells, with 18.28 stores on each road and 12.52 stores in each line cell on average. line cell on average.
The
The second
second largest
largest number
number comes
comes fromfrom 16,900
16,900 CPS,
CPS, which
which are are evenly
evenly located
located in in densely
densely populated
populated
areas. The CMS, however, are concentrated in the urban fringe due to
areas. The CMS, however, are concentrated in the urban fringe due to its demand for land useits demand for land use andand
the
sensitivity to the land lease. Therefore, the density of CPS is generally lower than
the sensitivity to the land lease. Therefore, the density of CPS is generally lower than that of the CMS. that of the CMS.
A 2 km bandwidth is used to calculate the NKDE value for all store types. The conditional
cumulative distribution function
Table (CCDF)summary
1. Descriptive is used toof calculate
retail storesthein statistical distribution of the NKDE
the study area.
value. The CCDF indicator, which is defined as F(α) = P(x > α), reflects the statistical distribution of
Store Road Average Stores Line Cell Average Stores Per
Store Type
samples. Store Subtypecan be derived when the CCDF value is larger than α. Generally,
The sum of probabilities Count Count Per Street Count Line Cell
the NKDE value ofOffice all store
supplytypes
stores follows
(OSS) a power-law
4285 897 distribution
4.7770 (Figure 3), referring3.1186
1374 to the fact
that the number of roads decreases geometrically with the increase of the NKDE value. Among all
Apparel shop stores (AOS) 15456 1245 12.4145 1829 8.4505
Home appliance stores (HAS)
storeSpecialty
types, the SS has the longest tail, which4919 1035
indicates 4.7527
that it involves more 1626
roads due 3.0252
to its dense
Drug stores (DGS) 3142 979 3.2094 1555 2.0206
stores On the contrary, the DS has the shortest tail. A similar power-law distribution has been
distribution. Car stores (CAS) 3811 434 8.7811 795 4.7937
(SS)
observed in seven SSCosmetics
subtypes, among
stores (CTS)which the3569AOS has860 the longest
4.1500 tail, and1367
the DGS has 2.6108
the shortest
tail (Figure 4). Other stores (OTS) 6587 1491 4.4178 2156 3.0552
Subtotal 41769 2285 18.2796 3335 12.5244
Department stores (DS) 249 121 2.0579 179 1.3911
Supermarkets (SMS) 2377 823 2.8882 1375 1.7287
Furniture stores (FS) 4211 492 8.5589 803 5.2441
Construction material stores (CMS) 13271 1147 11.5702 1887 7.0329
Consumer product stores (CPS) 16900 1966 8.5961 2967 5.6960
All Stores (AS) 78777 2652 29.7048 3934 20.0247
the NKDE value of all store types follows a power-law distribution (Figure 3), referring to the fact
that the number of roads decreases geometrically with the increase of the NKDE value. Among all
store types, the SS has the longest tail, which indicates that it involves more roads due to its dense
distribution. On the contrary, the DS has the shortest tail. A similar power-law distribution has been
observed in
Sustainability seven
2019, SS subtypes, among which the AOS has the longest tail, and the DGS has
11, 4539 8 ofthe
20
shortest tail (Figure 4).

Figure
Figure 3. The
3. The conditional
conditional cumulative
cumulative distribution
distribution function
function (CCDF)
(CCDF) ofnetwork-based
of the the network-based
kernelkernel
density
density estimation
estimation (NKDE) value for six store types. (NKDE) value for six store types.

Figure 4.
Figure 4. CCDF of the NKDE value
value for
for seven
seven SS
SS subtypes.
subtypes.

4.2. Network-Based Spatial

4.2. Network-Based Spatial Patterns
Patterns of
of Retail
Retail Stores
Stores
The method described
The method describedininSection
Section2.1 2.1is isused
usedto to calculate
calculate thethe density
density valuevalue of each
of each line line cell.
cell. The
The calculation
calculation is based
is based on each
on each lineline cell.
cell. It retrieves
It retrieves thethe pointevents
point eventsononall
allline
linecells
cells within
within aa given
given
bandwidth
bandwidth range
range(measured
(measured bybythethe
network
network distance) and calculates
distance) the corresponding
and calculates kernel density
the corresponding kernel
value
densitythrough the kernel
value through the function. Then, Then,
kernel function. the linethecell
linelength is used
cell length as the
is used asweight to derive
the weight the
to derive
line density, as shown in Figures 5–7. In the figures, the NKDE value of a store type
the line density, as shown in Figure 5–7. In the figures, the NKDE value of a store type represents the represents the
concentration of the stores on the road. The NKDE is visually divided into six classes based on the
Jenks natural breaks. All stores in the study area are mainly distributed between West 3rd Ring road,
Lianyungang-Khorgos Highway, Zhongzhou Avenue, and South 3rd Ring Road (Figure 5). They are
concentrated most densely in the Erqi Square near the Zhengzhou Railway Station (position O 1 in
Figure 5). As a traditional business center of Zhengzhou city, this area embodies many large shopping
malls. A large number of furniture and home appliance stores are located in the area adjacent to
Weilai Road, Zhongzhou Avenue, Zhengbian Road, and Longhai Expressway (position O 2 in Figure 5).
Additionally, many stores are concentrated in the area to the east of Huayuan Road, West of Zhongzhou
Avenue, and North of Sanquan Road (position O 3 in Figure 5), and at the intersection of Zhongzhou
concentrated most densely in the Erqi Square near the Zhengzhou Railway Station (position ① in
Figure 5). As a traditional business center of Zhengzhou city, this area embodies many large shopping
malls. A large number of furniture and home appliance stores are located in the area adjacent to
Weilai Road, Zhongzhou Avenue, Zhengbian Road, and Longhai Expressway (position ② in Figure
5). Additionally,
Sustainability 2019,many
11, 4539stores are concentrated in the area to the east of Huayuan Road,9 of West
20 of

Zhongzhou Avenue, and North of Sanquan Road (position ③ in Figure 5), and at the intersection of
Zhongzhou
Avenue and Avenue
South and South
3rd Ring Road3rd Ring Road
(position O
4 in Figure 5). We④
(position inperformed
also Figure 5).theWe also performed
network-based KDE the
network-based
for the six storeKDEtypes
for the six store
(Figure 6) andtypes (Figure
the seven 6) and the
SS subtypes seven
(Figure 7).SS subtypes
Table 2 shows(Figure 7). Table 2
the statistical
shows the statistical
results of the roadresults of thefor
KDE values road KDE
these values for these stores.
stores.

Figure 5. 5.
Figure Network-based
Network-basedKDE valuefor
KDE value forall
allretail
retail stores.
stores.

Table 2. Statistics of network-based KDE values by store type.

Store Type Store Subtype Average Median Maximum Standard Deviation

Office supply stores (OSS) 6.2003 3.2650 115.6580 8.5152
Apparel shop stores (AOS) 17.6005 6.6483 479.3170 29.8941
Home appliance stores (HAS) 6.1540 3.5878 129.5320 9.3261
Specialty stores Drug stores (DGS) 3.8343 2.2271 29.3573 4.1453
(SS) Car stores (CAS) 4.0902 1.1406 328.1910 16.7913
Cosmetics stores (CTS) 4.7676 3.1391 70.3148 5.9746
Other stores (OTS) 8.2742 5.9976 139.8360 8.4355
Subtotal 40.3303 22.3597 766.2540 52.1403
Department stores (DS) 0.3422 0.0388 11.8333 0.8177
Supermarkets (SMS) 3.1721 2.3872 63.3645 3.6940
Furniture stores (FS) 5.5538 1.5317 191.4990 13.9360
Construction material stores (CMS) 14.4089 5.1197 506.5090 35.1740
Consumer product stores (CPS) 17.9092 12.5695 406.3060 20.4603
All Stores (AS) 74.5491 45.1345 1105.7300 88.6215

(a) (b)
Sustainability 2019, 11, 4539 10 of 20
Figure 5. Network-based KDE value for all retail stores.

Sustainability 2019, 11, 4539 10 of 20

(a) (b)

(c) (d)

(e) (f)

Figure
Figure 6.
6. Network-based KDEvalue
Network-based KDE valuefor
forsix
sixstore
store types.
types. (a)(a)
SS;SS;
(b)(b)
DS;DS; (c) SMS;
(c) SMS; (d) (e)
(d) FS; FS;CMS;
(e) CMS; (f)
(f) CPS.
CPS.
 (e) (f)
SustainabilityFigure
2019, 11, 4539
6. Network-based KDE value for six store types. (a) SS; (b) DS; (c) SMS; (d) FS; (e) CMS; (f) 11 of 20
CPS.

Sustainability 2019, 11, 4539 11 of 20

(a) (b)

(c) (d)

(e) (f)

Figure
Figure 7. Network-based
7. Network-based KDEvalue
KDE valueof
of six
six SS
SSsubtypes.
subtypes.(a)(a)
OSS; (b) AOS;
OSS; (c) HAS;
(b) AOS; (d) DGS;
(c) HAS; (d)(e) CAS;(e)
DGS; (f) CAS;
CTS.
(f) CTS.

Table 2.and
4.3. Correlations between NKDE Statistics of network-based
Closeness Centrality KDE values by store type.
Store Type Store Subtype Average Median Maximum Standard Deviation
We calculated the GCC, LCC, and WCC values
Office supply stores (OSS) 6.2003
in the study115.6580
3.2650
area (Figure 8.5152
8). The definition of
the neighborhood in the LCC calculation
Apparel is the road
shop stores (AOS) 17.6005network
6.6483 community
479.3170 generated
29.8941 by the Combo
Home appliance stores (HAS) 6.1540 3.5878 129.5320 9.3261
algorithm (Figure 2). The closeness centrality manifests a significant level of spatial heterogeneity.
Specialty stores Drug stores (DGS) 3.8343 2.2271 29.3573 4.1453
The higher the closeness
(SS) centrality value
Car stores (CAS) of a road is,
4.0902 the shorter
1.1406 the road’s
328.1910 average
16.7913distance to other
roads. The GCC represents the average road distance, which gradually decreases from the city center
Cosmetics stores (CTS) 4.7676 3.1391 70.3148 5.9746
Other stores (OTS) 8.2742 5.9976 139.8360 8.4355
Subtotal 40.3303 22.3597 766.2540 52.1403
Department stores (DS) 0.3422 0.0388 11.8333 0.8177
Supermarkets (SMS) 3.1721 2.3872 63.3645 3.6940
Furniture stores (FS) 5.5538 1.5317 191.4990 13.9360
Construction material stores (CMS) 14.4089 5.1197 506.5090 35.1740
Consumer product stores (CPS) 17.9092 12.5695 406.3060 20.4603
Sustainability 2019, 11, 4539 12 of 20
Sustainability 2019, 11, 4539 12 of 20
The higher the closeness centrality value of a road is, the shorter the road’s average distance to other
roads. The GCC represents the average road distance, which gradually decreases from the city center
to the periphery (Figure 8a). However, the LCC shows the average distance in a subnetwork and
to the periphery (Figure 8a). However, the LCC shows the average distance in a subnetwork and
identifies a road with high centrality in a local area. Unlike GCC, there are many local areas with a
identifies a road with high centrality in a local area. Unlike GCC, there are many local areas with a
high LCC value. The old town area surrounding the Zhengzhou East Railway Station has the highest
high LCC value. The old town area surrounding the Zhengzhou East Railway Station has the highest
value (position O 1 in Figure 8b). The roads near the Zhengzhou East Railway Station (position O 2 in
value (position ① in Figure 8b). The roads near the Zhengzhou East Railway Station (position ②
Figure 8b) and the roads between the Zhengping Highway and Daxue South Road (position O 3 in
in Figure 8b) and the roads between the Zhengping Highway and Daxue South Road (position ③ in
Figure 8b) also have high LCC values. The WCC shows both the average distance of the global road
Figure 8b) also have high LCC values. The WCC shows both the average distance of the global road
network and local subnetworks. Roads with a high WCC value are mainly distributed near the railway
network and local subnetworks. Roads with a high WCC value are mainly distributed near the
station in the old town area. The roads near Erqi Square have the highest WCC value, which means
railway station in the old town area. The roads near Erqi Square have the highest WCC value, which
that those roads have a considerable level of accessibility to all other roads in the network.
means that those roads have a considerable level of accessibility to all other roads in the network.

(a)

(b)

Figure 8. Cont.
Sustainability 2019, 11, 4539 13 of 20
Sustainability 2019, 11, 4539 13 of 20

(c)

Figure 8. The closeness centrality of roads in the study area. (a) GCC; (b) LCC; (c) WCC.

The correlation
correlationcoefficient between
coefficient between the NKDE
the NKDEvalue value
and theand
closeness centralitycentrality
the closeness is also calculated,
is also
as shown inasTable
calculated, shown3. Among
in Tablethe six storethe
3. Among types, SS and
six store SMSSS
types, have
andtheSMShighest levelhighest
have the of correlation,
level of
meaning thatmeaning
correlation, they are relatively
that they accessible to allaccessible
are relatively other nodes toand
all have
otherconsiderable
nodes and havepotential to attract
considerable
customers. In particular, the high correlation coefficient is shown between SS and WCC,
potential to attract customers. In particular, the high correlation coefficient is shown between SS and between SS
and
WCC, GCC, as well
between SS as
andbetween
GCC, asSSwell
andasLCC. This SS
between is very
and LCC.likelyThis
due is
tovery
the location
likely dueadvantage of SS,
to the location
which are normally
advantage distributed
of SS, which in thedistributed
are normally center of the subnetworks.
in the However,
center of the the FSHowever,
subnetworks. and CMSthe show
FS
aand
lowCMS
levelshow
of correlation.
a low levelThis result is due
of correlation. to the
This relatively
result is due few stores
to the covering
relatively fewastores
large serve area:
covering a
Because of their
large serve area:relative
Becausesensitivity to land sensitivity
of their relative lease, thesetostores
land are generally
lease, located
these stores areingenerally
the urbanlocated
fringe.
in the urban fringe.
Table 3. The correlation coefficients between store NKDE and street closeness centrality.
Table 3. The correlation
Store Type
coefficients between store NKDE
Store Subtype GCC
and street closeness
LCC
centrality.
WCC
Store Type Office supply
Storestores (OSS)
Subtype 0.5602GCC
*** 0.5938
LCC*** 0.6309 ***
WCC
Apparel shop stores (AOS) 0.4702 *** 0.5446 *** 0.5657 ***
Office supply stores
Home appliance stores (HAS)
(OSS) 0.5602
0.4535 ***
*** 0.5938 ***
0.4105 ***
0.6309 ***
0.4676 ***
Specialty stores Apparel shop
Drug stores stores (AOS) 0.5666
(DGS) 0.4702
*** *** 0.5446******
0.3485 0.5657
0.4536***
***
(SS) Carappliance
stores (CAS)stores (HAS) 0.0173 # −0.0489 −0.0296
Home 0.4535 *** 0.4105****** 0.4676 ****
Cosmetics stores (CTS) 0.5172 *** 0.5204 *** 0.5623 ***
Specialty stores Drug stores (DGS)
Other stores (OTS)
0.5666 ***
0.6166 ***
0.3485 ***
0.4669 ***
0.4536 ***
0.5626 ***
(SS) Car stores (CAS)
Subtotal 0.0173
0.5391 *** # −0.0489
0.5284 ****** −0.0296 *
0.5790 ***
Cosmetics
Department stores (DS) stores (CTS) 0.5172
0.3367 *** *** 0.5204
0.5049 *** *** 0.5623 ***
0.4856 ***
Other stores (OTS)
Supermarkets (SMS) 0.6166
0.5344 *** *** 0.4669****** 0.5626
0.4564 0.5290***
***
Furniture stores (FS) 0.1731 *** 0.0755 *** 0.1236 ***
Subtotal 0.5391 *** 0.5284 *** 0.5790 ***
Construction material stores (CMS) 0.2027 *** 0.0728 *** 0.1334 ***
Department
Consumer stores
product stores (DS)
(CPS) 0.3367
0.4432 *** *** 0.5049****** 0.4856
0.2722 0.3568***
***
All Stores (AS) (SMS)
Supermarkets 0.5322 *** *** 0.4542
0.5344 0.4564 ****** 0.5265***
0.5290 ***
Note: Furniture stores
*** significance level(FS) 0.1731
of 0.001, * significance level of 0.1,*** 0.0755
and # not *** 0.1236 ***
significant.
Construction material stores (CMS) 0.2027 *** 0.0728 *** 0.1334 ***
To examineConsumer
the effects product
of different bandwidths
stores (CPS) in the NKDE,
0.4432 three
*** bandwidths,
0.2722 *** including
0.3568 *** 0.5-km,
1 km, and 2 km, were implemented.
All Stores (AS) Then, we compared 0.5322 the correlation
*** 0.4542 with*** GCC,0.5265
LCC, ***
and WCC
with these three bandwidths (Figure 9). The results show a positive correlation
Note: *** significance level of 0.001, * significance level of 0.1, and # not significant. coefficient for all
bandwidths except CAS; however, the correlation coefficient is comparatively small under a small
To examine the effects of different bandwidths in the NKDE, three bandwidths, including 0.5-
km, 1 km, and 2 km, were implemented. Then, we compared the correlation with GCC, LCC, and
Sustainability 2019, 11, 4539 14 of 20
Sustainability 2019, 11, 4539 14 of 20
WCC with these three bandwidths (Figure 9). The results show a positive correlation coefficient for
all bandwidths except CAS; however, the correlation coefficient is comparatively small under a small
bandwidth. For example,
bandwidth. For example, the
the correlation
correlation coefficient
coefficient between
between SSSS and
and WCC
WCC decreases
decreases from
from 0.58
0.58 to
to 0.48,
0.48,
and that between OSS and WCC decreases from 0.63 to 0.48. However, the difference
and that between OSS and WCC decreases from 0.63 to 0.48. However, the difference betweenbetween different
bandwidths is not significant.
different bandwidths is not significant.

Figure 9. Correlations
Correlations between the NKDE value and the the closeness
closeness centrality
centrality with different bandwidth.
* and ** represent a significance level of 0.1 and 0.01, respectively; # represents non-significance; other
significance levels are at 0.001.

4.4. Discussion
4.4. Discussion
Spatial distribution and
Spatial distribution and location
location analysis
analysis of of business
business retailers
retailers havehave beenbeen the the focus
focus of of urban
urban
planning and business administration. Most existing studies used regular
planning and business administration. Most existing studies used regular grid cells or administrative grid cells or administrative
units
units toto perform
perform spatial
spatial analysis. These analyses
analysis. These analyses suffer
suffer considerably
considerably from from statistical
statistical biases,
biases, such
such asas
the
the MAUP and the edge effect. These methodological caveats are addressed in the paper using aa
MAUP and the edge effect. These methodological caveats are addressed in the paper using
network-based
network-based analysis.analysis. The The correlation
correlation analysis
analysis between
between the the store
store density
density and and thethe street
street closeness
closeness
centrality are also discussed. The results show various levels of
centrality are also discussed. The results show various levels of correlation by store type. correlation by store type.
Specifically,
Specifically, in in aa road
road network,
network, the the node
node withwith the
the highest
highest level
level ofof closeness
closeness centrality
centrality value
value will
will
always
always stay close to
stay close to the
the geometric
geometric center
center of of the
the network
network [62,63];
[62,63]; the higher the
the higher the closeness
closeness centrality
centrality of of
a road is, the shorter its average distance to other roads and the better
a road is, the shorter its average distance to other roads and the better its accessibility. In the study its accessibility. In the study
area,
area, the
the index
index of of GCC
GCC shows
shows aa circular
circular distribution
distribution similar
similar to to concentric
concentric circles
circles in in space,
space, while
while LCCLCC
shows
shows a a similar
similar distribution
distribution in in each
each subnetwork
subnetwork (Figure (Figure 8). 8). We
We cancan see
see that
that thethe closeness
closeness centrality
centrality
represents
represents the the distance
distance between
between aa node node andand the the centroid
centroid of of the
the network:
network: The The higher
higher the the closeness
closeness
centrality is, the closer the node to the centroid of the network.
centrality is, the closer the node to the centroid of the network. As nodes in a network As nodes in a network do not distribute
do not
evenly, the centroid does not entirely coincide with the geometric
distribute evenly, the centroid does not entirely coincide with the geometric center of the network.center of the network. Therefore,
in actual road
Therefore, networks,
in actual road the centroids
networks, thereflected
centroidsbyreflected
the closeness
by thecentrality
closenesscan characterize
centrality the kernel
can characterize
of
thea kernel
city or the of anetwork-based centroid. The centroid.
city or the network-based irregular concentric
The irregular spatial circles inspatial
concentric the distribution
circles in the are
able to indicate the city structure kernel. Furthermore, stores with
distribution are able to indicate the city structure kernel. Furthermore, stores with a high LCC a high LCC correlation tend to
distribute in the center of the subnetwork. In reality, such stores tend
correlation tend to distribute in the center of the subnetwork. In reality, such stores tend to have to have relative accessibility and
are able to
relative attract passing
accessibility and traffic
are able andtocustomers,
attract passingthus helping to gain
traffic and a business
customers, potential
thus helping and toachieve
gain a
sustainable development.
business potential and achieve sustainable development.
The methods and
The methods and findings
findingsofofthe thepaper
papercan can provide
provide valuable
valuable insights
insights intointourbanurban planning.
planning. By
By analyzing
analyzing the the correlation
correlation between
between the facility
the facility distribution
distribution and theandcloseness
the closeness centrality
centrality of urban
of urban road
road networks,
networks, it is it is possible
possible to optimize
to optimize thethe spatial
spatial patterns
patterns ofofthethecommercial
commercialfacilities.
facilities. ForFor example,
example,
results drawn from
results drawn from thethestudy
studysuggest
suggestthat thatfuture
futureplanning
planningstrategies
strategies could
could include
include relocating
relocating part
part of
of the facilities, such as those with a high demand for space but low
the facilities, such as those with a high demand for space but low sensitivity to travel distance, to the sensitivity to travel distance,
to the urban
urban fringe.fringe. In addition,
In addition, the identification
the identification of aofcity’s
a city’s sub-centers
sub-centers willwillhelp
helptotodelineate
delineate the the city’s
city’s
polycentric
polycentric spatial
spatial structure,
structure, providing
providing evidence
evidence for for taking
taking future
future planning
planninginitiatives.
initiatives.
Physical accessibility is a crucial factor for the location strategy
Physical accessibility is a crucial factor for the location strategy of retailers. The of retailers. The results
results of of the
the
correlation analysis using the network struture is consistent with similar
correlation analysis using the network struture is consistent with similar analyses in other case analyses in other case studies,
including BolognaBologna
studies, including [13], Barcelona [14], Changchun
[13], Barcelona [14], Changchun[15], and[15],Guangzhou
and Guangzhou [16], all[16],
of which
all of used
which a
grid cell structure. These studies confirm the hypothesis that street
used a grid cell structure. These studies confirm the hypothesis that street centrality is a crucial centrality is a crucial feature in
Sustainability 2019, 11, 4539 15 of 20

the urban structure and land-use pattern and also indicate that commercial retail market tends to be
concentrated in areas with better centrality. Results derived from the store subtypes could be compared
with these studies and are significant for capturing trends and strategies of urban development in
different cities. For example, in the study area, the correlation between CPS and GCC is higher than that
between DS and GCC. In Changchun, a northeastern Chinese city, the conclusion is the opposite [15].
This observation is mainly attributed to the development strategy of the Zhengdong New District
in the study area. The new development in the district is separate from expanding the old town
center. It aims to establish a suburban center by following the future development goals outlined in
the planning blueprint of Zhengzhou City 2010–2030. This development plan has encouraged the
establishment of large shopping centers (i.e., DS in the analysis) in the new suburban center and can be
identified by the global correlation coefficient (e.g., between DS and GCC). It can be observed from the
analysis that urban planning and development policies differ across cities and could largely influence
the spatial structure of the retail industry. These patterns could be identified using the network-based
analysis as exemplified in the paper.
The correlation difference between the network-based and cell-based KDE (CKDE) are compared.
Firstly, the CKDE for various commercial POI and three closeness centralities are calculated based on
100 m grid cells, and then correlation analysis is conducted (Table 4, Figures 10–12). The bandwidth of
the CKDE is set to 2 km. It is illustrated that the correlation coefficients with the two types of the KDE
are generally consistent in the distribution trends for different store types, but the correlation coefficient
based on the CKDE is higher than that of the NKDE. This result is attributed to the smoothing effect of
the CKDE as the small difference in density values between adjacent cells in the CKDE calculation. In
fact, since retail stores has most points distributed along the street, the NKDE better reveals its spatial
distribution pattern.

Table 4. The correlation coefficients of cell-based KDE between stores and closeness centrality.

Store Type GCC LCC WCC

OSS 0.7772 *** 0.8390 *** 0.8461 ***
AOS 0.6487 *** 0.7850 *** 0.7591 ***
HAS 0.6842 *** 0.7001 *** 0.7339 ***
DGS 0.7556 *** 0.6845 *** 0.7439 ***
SS
CAS 0.0458 *** -0.0252 *** 0.0068 #
CTS 0.7370 *** 0.7709 *** 0.7865 ***
OTS 0.8251 *** 0.7482 *** 0.8187 ***
Subtotal 0.7381 *** 0.7914 *** 0.8041 ***
DS 0.6354 *** 0.7740 *** 0.7395 ***
SMS 0.8138 *** 0.7650 *** 0.8283 ***
FS 0.3720 *** 0.2710 *** 0.3429 ***
CMS 0.3325 *** 0.2206 *** 0.2954 ***
CPS 0.7055 *** 0.6189 *** 0.6874 ***
AS 0.7383 *** 0.7128 *** 0.7630 ***
Note: *** significance level of 0.001, # not significant.

This study also has limitations. First, there are methodological caveats in the analysis. For example,
the paper employs the closeness centrality of the network to measure the network structure while
ignoring other network metrics. In the process of closeness centrality calculation, the shortest path
algorithm only takes into account the distance metric and fails to consider the hierarchical structure of
the road network, such as the speed limit. In the calculation of the KDE, the bandwidth parameter
should be optimized using fitted models. Second, the identification of the spatial pattern is only
focused on the supply side (i.e., POI data) while ignoring the demand or the consumers’ need. Future
analysis of the retailer accessibility can employ spatial interaction models, such as the two-step floating
catchment area method [18]. Third, the distribution of commercial POI is influenced by factors separate
from the transport infrastructure, such as commuting population and land use. Thus, it is necessary to
Sustainability 2019, 11, 4539 16 of 20

weight in these factors using other spatial measures such as the geographically weighted regressing
method. Lastly, the study only considers the brick-and-mortar stores and disregards other retail forms,
especially online retailers. How to evaluate the impact of the internet retailer on urban development
with cutting edge spatial methods would be an area worthy of further exploration.
Sustainability 2019, 11, 4539 16 of 20

GCC
0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0
SS
S
AS

S
S
SS

S
FS

S
DS

S
AS

TS

SM

CM
CT
AO

DG

CP

A
O

O
C

Cell-based KDE Network-based KDE

Figure 10. The

The comparison
comparison of
of correlation
correlation coefficients with GCC between CKDE and NKDE.

LCC
0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0
S
FS

S
DS
S
AS

SS
SS

S
AS
S
S
TS

SM

CP
AO

DG
CA
CT

CM

-0.1
O

Cell-based KDE Network-based KDE

Figure
Figure 11.
11. The
The comparison
comparison of
of correlation
correlation coefficients
coefficients with
with LCC
LCC between CKDE and
between CKDE and NKDE.
NKDE.
Sustainability 2019, 11, 4539 17 of 20
Sustainability 2019, 11, 4539 17 of 20

WCC
0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0.0
S
AS

S
SS

FS
SS

S
S
DS

AS
S
S
TS

SM
AO

DG

CP
CA
CT

CM
-0.1
O

Cell-based KDE Network-based KDE

Figure
Figure 12. The
The comparison
comparison of
of correlation
correlation coefficients
coefficients with WCC between CKDE and NKDE.

5. Conclusions
This study also has limitations. First, there are methodological caveats in the analysis. For
example, the paper
Identifying theemploys the closeness
spatial pattern centrality
of urban of the
retailers network to
is valuable formeasure the network
understanding landstructure
use and
while
businessignoring other
potential of anetwork metrics.
city. In this paper,In the
the store
process of closeness
density centrality
is modeled using acalculation, the shortest
network structure and
path algorithm only takes into account the distance metric and fails to
quantified by network metrics. Correlations between the store density and the network closeness consider the hierarchical
structure
centrality of
arethe road network,
discussed based onsuchbothasthetheglobal
speedand
limit.
localIneffects
the calculation of thestore
across different KDE,types.
the bandwidth
The paper
parameter
also derives should be optimized
the store density byusingstorefitted models.
type and Second,
subtype thethe
using identification
NKDE method. of theThree
spatial pattern
closeness
is only focused on the supply side (i.e., POI data) while ignoring the demand
centrality indicators (i.e., GCC, LCC, and WCC) are proposed to gauge the spatial pattern and or the consumers’ need.
are
Future analysis of the retailer accessibility can employ spatial interaction
correlated with the NKDE. Further analysis indicates that stores with a high correlation tend to models, such as the two-
be
step floating catchment area method [18]. Third, the distribution of commercial POI is
distributed in either the city center (i.e., GCC) or the subnetwork center (i.e., LCC). Retail stores located influenced by
factors
on these separate fromathe
roads have transport
relative infrastructure,
location advantage andsuchhave
as commuting population
the potential to attractand land traffic
passing use. Thus,
and
it is necessary
customers. to weight
These findings in can
these factorssuggestions
provide using othertospatial measures
optimize such pattern
the spatial as the geographically
of commercial
weighted regressing
facilities and method.
are valuable Lastly, the
for promoting thestudy only considers
sustainable development the ofbrick-and-mortar
the city. stores and
disregards other retail forms, especially online retailers. How to evaluate the impact of the internet
Author Contributions:
retailer Z.H. designed
on urban development withand performed
cutting the experiments
edge spatial methodsand wrote
would bethis
an paper. C.C. designed
area worthy the
of further
experiments and analyzed the data. C.M. supervised the implementation of the project. H.W. contributed data
exploration.
collection and analysis. X.C. edited the manuscript.
Funding:
5. This research was funded by National Natural Science Foundation of China, grant number 41601116 and
Conclusions
41871316; Key Project of Humanities and Social Sciences by Key Research Base of the Ministry of Education, China,
GrantIdentifying
number 16JJD770019; Science
the spatial and Technology
pattern Foundation
of urban retailers is of Henan Province,
valuable grant numberland
for understanding 172102110286;
use and
and the Henan University Science Foundation for Young Talents.
business potential of a city. In this paper, the store density is modeled using a network structure and
Acknowledgments:
quantified by network Special thanks
metrics. go to the editor
Correlations betweenandtheanonymous reviewers
store density and theof this papercloseness
network for their
constructive comments.
centrality are discussed based on both the global and local effects across different store types. The
Conflicts of Interest: The authors declare no conflict of interest.
paper also derives the store density by store type and subtype using the NKDE method. Three
closeness centrality indicators (i.e., GCC, LCC, and WCC) are proposed to gauge the spatial pattern
References
and are correlated with the NKDE. Further analysis indicates that stores with a high correlation tend
to
1. be Barry,
distributed in either
R.B.; Joel, the cityM.C.
R.E.; Patrali, center (i.e.,
Retail GCC) or the
Management: subnetwork
A Strategic center
Approach, (i.e.,
13th ed.;LCC). Retail
Pearson stores
Education:
located on these
Harlow, roadspp.
UK, 2018; have a relative location advantage and have the potential to attract passing
22–42.
2. Glaeser,
traffic E.L.; Kolko, These
and customers. J.; Saiz, findings
A. Consumer
cancity. J. Econ.suggestions
provide Geogr. 2001, 1,
to27–50. [CrossRef]
optimize the spatial pattern of
3.
commercial China’s New
Wang, S.facilities Retail
and are Economy:
valuable forApromoting
Geographic the
Perspective; Routledge:
sustainable New York,
development NY,city.
of the USA, 2014;
pp. 10–60.
Author Contributions: Z.H. designed and performed the experiments and wrote this paper. C.C. designed the
experiments and analyzed the data. C.M. supervised the implementation of the project. H.W. contributed data
collection and analysis. X. C. edited the manuscript.
Sustainability 2019, 11, 4539 18 of 20

4. Goldman, A. The transfer of retail formats into developing economies: The example of China. J. Retail. 2001,
77, 221–242. [CrossRef]
5. Thomas, W.; Susan, G.M. Retail gravitation and economic impact: A market-driven analytical framework for
bike-share station location analysis in the United States. Int. J. Sustain. Transp. 2016, 10, 247–259.
6. Les, D.; Michalis, P.; Alex, S. Estimating attractiveness, hierarchy and catchment area extents for a national
set of retail centre agglomerations. J. Retail. Consum. Serv. 2016, 28, 78–90.
7. Michael, F.G. LACS: A Location-Allocation Mode for Retail Site Selection. J. Retail. 1984, 60, 84–100.
8. Norat, R.T.; Amparo, B.P.; Juan, B.V.; Francisco, M.V. The retail site location decision process using GIS and
the analytical hierarchy process. Appl. Geogr. 2013, 40, 191–198.
9. Mike, G. The Poetics of Cities: Designing Neighborhoods That Work; The Ohio State University Press: Columbus,
OH, USA, 1995; pp. 1–16.
10. Michael, F.D. The geometry of central place theory. Geogr. Ann. B 1965, 47, 111–124.
11. Gordon, F.M. Agglomeration and central place theory: A review of the literature. Int. Reg. Sci. Rev. 1984, 9,
1–42.
12. Guy, C. The Retail Development Process. Location, Property and Planning; Routeledge: London, UK, 1994;
pp. 12–36.
13. Sergio, P.; Emanuele, S.; Valentino, I.; Roberto, M.; Vito, L.; Alessio, C.; Wang, F.H.; Salvatore, S. Street
centrality and densities of retail and services in Bologna, Italy. Environ. Plan. B 2009, 36, 450–465.
14. Sergio, P.; Vito, L.; Wang, F.H.; Salvador, R.; Emanuele, S.; Salvatore, S.; Alessio, C.; Eugenio, B.; Francisco, C.;
Berta, C. Street centrality and the location of economic activities in Barcelona. Urban Stud. 2012, 49, 1471–1488.
15. Wang, F.H.; Chen, C.; Xiu, C.L.; Zhang, P.Y. Location analysis of retail stores in Changchun, China: A street
centrality perspective. Cities 2014, 41, 54–63. [CrossRef]
16. Lin, G.; Chen, X.X.; Liang, Y.T. The location of retail stores and street centrality in Guangzhou, China.
Appl. Geogr. 2018, 100, 12–20. [CrossRef]
17. He, S.W.; Yu, S.; Wei, P.; Fang, C.L. A spatial design network analysis of street networks and the locations
of leisure entertainment activities: A case study of Wuhan, China. Sustain. Cities Soc. 2019, 44, 880–887.
[CrossRef]
18. Chen, X. Take the edge off: A hybrid geographic food access measure. Appl. Geogr. 2017, 87, 149–159.
[CrossRef]
19. Matthew, C.; Steve Ti Tim, H.; Taner, O. Public Places Urban Spaces, The Dimensions of Urban Design, 2nd ed.;
Routledge: New York, NY, USA, 2012; pp. 180–190.
20. Jacobs, J. The Death and Life of Great American Cities: The Failure of Town Planning; Random House: New York,
NY, USA, 1961; pp. 1–25.
21. Stephen, M. Streets and Patterns; Routledge: New York, NY, USA, 2005; pp. 235–258.
22. Shen, Y.; Kayvan, K. Urban function connectivity: Characterisation of functional urban streets with social
media check-in data. Cities 2016, 55, 9–21. [CrossRef]
23. Okabe, A.; Okunuki, K. A computational method for estimating the demand of retail stores on a street
network and its implementation in GIS. Trans. GIS 2001, 5, 209–220. [CrossRef]
24. Hillier, B.; Hanson, J. The Social Logic of Space; Cambridge University Press: London, UK, 1989; pp. 1–32.
25. Hillier, B. Space Is the Machine: A Configurational Theory of Architecture; Space Syntax: London, UK, 2007;
pp. 21–52.
26. Itzhak, O.; Ran, G. Spatial patterns of retail activity and street network structure in new and traditional
Israeli cities. Urban Geogr. 2016, 37, 629–649.
27. Wang, F.H.; Antipova, A.; Sergio, P. Street centrality and land use intensity in Baton Rouge, Louisiana.
J. Transp. Geogr. 2011, 19, 285–293. [CrossRef]
28. Wang, S.; Xu, G.; Guo, Q.S. Street Centralities and Land Use Intensities Based on Points of Interest (POI) in
Shenzhen, China. ISPRS. Int. J. Geo-Inf. 2018, 7, 425. [CrossRef]
29. Kevin, L. The Image of the City; MIT Press: Boston, MA, USA, 1960; pp. 1–26.
30. Ian, S.; Nikhil, N.; Carlo, R.; Raphäel, P. Green streets− Quantifying and mapping urban trees with street-level
imagery and computer vision. Landsc. Urban Plan. 2017, 165, 93–101.
31. Li, X.J.; Zhang, C.R.; Li, W.D.; Robert, R.; Meng, Q.Y.; Zhang, W.X. Assessing street-level urban greenery
using Google Street View and a modified green view index. Urban For. Urban Green. 2015, 14, 675–685.
[CrossRef]
Sustainability 2019, 11, 4539 19 of 20

32. Long, Y.; Liu, L. How green are the streets? An analysis for central areas of Chinese cities using Tencent
Street View. PLoS ONE 2017, 12, e0171110.
33. Gong, F.Y.; Zeng, Z.C.; Zhang, F.; Li, X.J.; Edward, N.; Leslie, K.N. Mapping sky, tree, and building view
factors of street canyons in a high-density urban environment. Build. Environ. 2018, 134, 155–167. [CrossRef]
34. Yin, Y.; Wang, Z.X. Measuring visual enclosure for street walkability: Using machine learning algorithms
and Google Street View imagery. Appl. Geogr. 2016, 76, 147–153. [CrossRef]
35. Li, X.J.; Carlo, R.; Ian, S. Quantifying the shade provision of street trees in urban landscape: A case study in
Boston, USA, using Google Street View. Landsc. Urban Plan. 2018, 169, 81–91. [CrossRef]
36. Li, X.J.; Carlo, R. Mapping the spatio-temporal distribution of solar radiation within street canyons of Boston
using Google Street View panoramas and building height model. Landsc. Urban Plan. 2018. [CrossRef]
37. Hao, J.W.; Zhu, J.; Zhong, R. The rise of big data on urban studies and planning practices in China: Review and
open research issues. J. Urban Manag. 2015, 4, 92–124. [CrossRef]
38. Zhu, D.; Wang, N.H.; Wu, L.; Liu, Y. Street as a big geo-data assembly and analysis unit in urban studies:
A case study using Beijing taxi data. Appl. Geogr. 2017, 86, 152–164. [CrossRef]
39. Wu, L.; Zhi, Y.; Sui, Z.W.; Liu, Y. Intra-urban human mobility and activity transition: Evidence from social
media check-in data. PLoS ONE 2014, 9, e97010. [CrossRef]
40. Pucci, P.; Fabio, M.; Paolo, T. Mapping Urban Practices through Mobile Phone Data; Springer: Berlin, Germany,
2015; pp. 21–70.
41. Yuan, J.; Zheng, Y.; Xie, X. Discovering regions of different functions in a city using human mobility and
POIs. In Proceedings of the 18th ACM SIGKDD International Conference on Knowledge Discovery and
Data Mining, Beijing, China, 12–16 August 2012; pp. 1–9.
42. Gao, S.; Janowicz, K.; Couclelis, H. Extracting urban functional regions from points of interest and human
activities on location-based social networks. Trans. GIS 2017, 21, 446–467. [CrossRef]
43. Hu, T.Y.; Yang, J.; Li, X.C.; Gong, P. Mapping urban land use by using landsat images and open social data.
Remote Sens. 2016, 8, 151. [CrossRef]
44. Liu, X.P.; He, J.L.; Yao, Y.; Zhang, J.B.; Liang, H.L.; Wang, H.; Hong, Y. Classifying urban land use by
integrating remote sensing and social media data. Int. J. Geogr. Inf. Sci. 2017, 31, 1675–1696. [CrossRef]
45. Biłozor, A.; Kowalczyk, A.M.; Bajerowski, T. Theory of Scale-Free Networks as a New Tool in Researching
the Structure and Optimization of Spatial Planning. J. Urban Plan. Dev. 2018, 144, 04018005. [CrossRef]
46. Kowalczyk, A.M. The analysis of networks space structures as important elements of sustainable space
development. In Proceedings of the International Conference on Environmental Engineering, Saulėtekio,
Lithuania, 27–28 April 2017; pp. 1–6.
47. Xie, Z.X.; Yan, J. Kernel density estimation of traffic accidents in a network space. Comput. Environ. Urban
Syst. 2008, 32, 396–406. [CrossRef]
48. Christopher, L. Spatial Data Analysis: An Introduction for GIS Users; Oxford University Press: London, UK,
2010; pp. 102–169.
49. Tang, L.L.; Kan, Z.H.; Zhang, X.; Sun, F.; Xue, Y.; Li, Q.Q. A network Kernel Density Estimation for linear
features in space–time analysis of big trace data. Int. J. Geogr. Inf. Sci. 2016, 30, 1717–1737. [CrossRef]
50. Okabe, A.; Kokichi, S. Spatial Analysis along Networks: Statistical and Computational Methods; John Wiley &
Sons: Chichester, UK, 2012; pp. 171–190.
51. She, B.; Zhu, X.Y.; Ye, X.Y.; Guo, W.; Su, K.H.; Jay, L. Weighted network Voronoi Diagrams for local spatial
analysis. Comput. Environ. Urban Syst. 2015, 52, 70–80. [CrossRef]
52. Stanley, W.; Katherine, F. Social Network Analysis: Methods and Applications; Cambridge University Press:
Cambridge, UK, 1994; pp. 1–35.
53. Linton, C.F. Centrality in social networks conceptual clarification. Soc. Netw. 1978, 1, 215–239.
54. Emanuele, S.; Matheus, V.; Luciano, F.C.; Alessio, C.; Sergio, P.; Vito, L. Urban street networks, a comparative
analysis of ten European cities. Environ. Plan. B 2013, 40, 1071–1086.
55. Sergio, P.; Paolo, C.; Vito, L. The network analysis of urban streets: A primal approach. Environ. Plan. B 2006,
33, 705–725.
56. Newman, M.E. Modularity and community structure in networks. Proc. Natl. Acad. Sci. USA 2006, 103,
8577–8582. [CrossRef]
57. Huang, L.P.; Yang, Y.J.; Gao, H.P.; Zhao, X.H.; Du, Z.W. Comparing Community Detection Algorithms in
Transport Networks via Points of Interest. IEEE Access 2018, 6, 29729–29738. [CrossRef]
Sustainability 2019, 11, 4539 20 of 20

58. Stanislav, S.; Riccardo, C.; Alexander, B.; Carlo, R. General optimization technique for high-quality community
detection in complex networks. Phys. Rev. 2014, 90, 1–8.
59. Ghalmane, Z.; El Hassouni, M.; Cherifi, C.; Cherifi, H. Centrality in modular networks. EPJ Data Sci. 2019, 8,
1–27. [CrossRef]
60. Andres, S.; Michael, M. Urban network analysis: A new toolbox for measuring city form in ArcGIS. In
Proceedings of the 2012 Symposium on Simulation for Architecture and Urban Design, Orlando, FL, USA,
26–29 March 2012; pp. 1–15.
61. Wang, X.F.; John, T. Zhengzhou—Political economy of an emerging Chinese megacity. Cities 2019, 84, 104–111.
[CrossRef]
62. Jiang, B. Street hierarchies: A minority of streets account for a majority of traffic flow. Int. J. Geogr. Inf. Sci.
2009, 23, 1033–1048. [CrossRef]
63. Alexander, E.; Michael, L.; Kay, W.A. Graph-theoretical analysis of the Swiss road and railway networks over
time. Netw. Spat. Econ. 2009, 9, 379–400.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).