EN中文

Journal of Geographical Sciences >

2018 , Vol. 28 >Issue 12: 1793 - 1809

DOI: https://doi.org/10.1007/s11442-018-1579-5

Research Articles

Hierarchy, clusters, and spatial differences in Chinese inter-city networks constructed by scientific collaborators

  • MA Haitao , 1 ,
  • FANG Chuanglin 1 ,
  • LIN Sainan 2 ,
  • HUANG Xiaodong 3 ,
  • XU Chengdong 1
Expand
  • 1. Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China
  • 2. School of Urban Design, Wuhan University, Wuhan 430072, China
  • 3. Key Research Institute of Yellow River Civilization and Sustainable Development, Henan University, Kaifeng 475001, Henan, China

Author: Ma Haitao (1979-), PhD and Associate Professor, specialized in urban network and innovation. E-mail:

Received date: 2017-09-17

  Accepted date: 2018-01-10

  Online published: 2018-12-20

Supported by

National Natural Science Foundation of China, No.41571151, No.41590842, No.71433008

Copyright

Journal of Geographical Sciences, All Rights Reserved
Fold

Abstract

The Chinese urban system is currently experiencing a fundamental shift, as it moves from a size-based hierarchy to a network-based system. Contemporary studies of city networks have tended to focus on economic interactions without paying sufficient attention to the issue of knowledge flow. Using data on co-authored papers obtained from China Academic Journal Network Publishing Database (CAJNPD) during 2014-2016, this study explores several features of the scientific collaboration network between Chinese mainland cities. The study concludes that: (1) the spatial organization of scientific cooperation amongst Chinese cities is shifting from a jurisdiction-based hierarchical system to a networked system; and (2) several highly intra-connected city regions were found to exist in the network of knowledge, and such regions had more average internal linkages (14.21) than external linkages (8.69), and higher average internal linkage degrees (14.43) than external linkage degrees (10.43); and (3) differences existed in terms of inter-region connectivity between the Western, Eastern, and Central China regional networks (the average INCD of the three regional networks were 109.65, 95.81, and 71.88). We suggest that China should engage in the development of regional and sub-regional scientific centers to achieve the goal of building an innovative country. Whilst findings reveal a high degree of concentration in those networks - a characteristic which reflects the hierarchical nature of China’s urban economic structure - the actual spatial distribution of city networks of knowledge flow was found to be different from that of city networks based on economic outputs or population.

Key words： Chinese city networks; knowledge flows; scientific collaboration; co-authored papers; knowledge city clusters; China academic journal; innovative country; regional differences

Cite this article

MA Haitao , FANG Chuanglin , LIN Sainan , HUANG Xiaodong , XU Chengdong . Hierarchy, clusters, and spatial differences in Chinese inter-city networks constructed by scientific collaborators[J]. Journal of Geographical Sciences, 2018 , 28(12) : 1793 -1809 . DOI: 10.1007/s11442-018-1579-5

1 Introduction

Inter-city relations change in response to developments in society, economy, and technology; as such, those relations constitute a crucial topic for urban researchers (Neal, 2013; Taylor and Derudder, 2017). With increased globalization and the development of a world economy, cities around the globe have become deeply interwoven. In light of this interconnectedness, the research community has shown increased interest in “networked” cities, locating them at the core content of inter-city relations. Moreover, since the beginning of the 21st century, the knowledge economy has become one of the most important features of the world economy (Fang et al., 2014); this development has seen more frequent and deeper knowledge flows between enterprises (Martin and Moodysson, 2013), research institutions (Leydesdorff and Persson, 2010), cities (Matthiessen et al., 2010; Ma et al., 2014; Li and Phelps, 2018), regions (Liang and Zhu, 2002) and nations. City networks are constructed around various flows, the importance of which changes during different development stages and time periods. In the contemporary era, knowledge flow has influenced inter-city relations profoundly and become a crucial aspect of city networks.
China is a developing country with a long-established hierarchical management system, which has strongly influenced the spatial organization of its cities (Wu, 2002; Ma, 2005). Despite the established nature of this urban system, as a result of the development of inter-city transportation systems, information infrastructure and the social economy, interactions are increasingly occurring between cities from different levels of the hierarchy and different regions. The Chinese urban system is thus experiencing a fundamental shift from a jurisdiction-based hierarchical system to a networked one (Lu and Huang, 2012; Derudder et al., 2013; Zhen et al., 2012). From the beginning of this century onward, inter-city knowledge communication and active innovative collaboration has also increased, owing to the government’s goal of building an innovation-oriented country. However, most recent research on Chinese city networks has focused either on flows of people or economic flows while little attention has been paid to the significant flow of knowledge (Li and Phelps, 2018). What are the spatial characteristics of Chinese city networks with respect to knowledge flow? How has the network system developed? What are the determining factors? Are there any differences between city networks based on knowledge flow and other flows? These are the research questions that motivate this study.

2 Literature review

City networks are a traditional research topic in the disciplines of human geography and urban studies. Whilst Harris and Ullman (1945) previously claimed that the internal structure of cities formed what could be termed “the nature of cities”, we argue that no city can be isolated from other cities. Rather, cities exist in a world of flows - population flows, material flows, and information transfers (Smith and Timberlake, 1995). Cities act as centers in relation to these flows, and within global, national, and regional economic systems, thus playing roles as major “nodes” (Friedmann, 1995; Wang et al., 2015). As such, the study of the relationships between cities and their inter-connectedness as spatial points (so-called “systems thinking”) is crucial to urban studies, an approach which Taylor (2004) references in his notion of “the second nature of cities”.
Early studies of inter-city relations tended to focus on “the urban system” (Zhou and Yang, 1987), considering issues of urban functional, administrative, and spatial structure. In studies influenced by central place theory, cities within a country have often been interpreted as forming a system that is dominated by the major cities and that follows a stepped, hierarchical structure. Most of these types of studies were carried out through examinations of the scale of urban population (Bourne, 1975). However, with breakthroughs in communications and information technology, and continuing economic globalization and regional integration, “spaces of flows” have gradually taken the place of “spaces of places” as the dominant conception of spatial form (Castells, 1989; Taylor et al., 2010). Rather than the scale status of a city, it is a city’s network status (Neal, 2011) and inter-city relations (Esparza and Krmenec, 2000) that have now become the focus of interest among scholars. At the end of the 20th century, research interests therefore shifted from the “urban system” to the “urban network” in the field of urban geography (Neal, 2011) - a new paradigm that better explains the organizational structure of urban space (Camagni and Capello, 2004).
Research on city networks has made considerable progress in the past decade. Whilst recent studies have made significant contributions to the development of theory and methodology in city network studies, a few research issues require further exploration. Firstly, it must be noted that increasing debates exist on the spatial scales of city networks. There is no doubt that Globalization and World Cities (GaWC) research reveals important features of city networks based on comparisons at different spatial scales. For examples, Taylor discovered the non-hierarchical urban structure of the world city network based on its comparison with the USA’s national urban system (Taylor et al., 2002). Taylor’ and Derudder found that some Chinese cities with high connectivity play an important role in the world city network based on their analysis of the global city network (Taylor, 2012; Derudder et al., 2013). However, the question remains: can these global-scale studies of city networks reflect all the features of national-scale city networks? Can national-scale city networks simply be seen as a sub-network of the world city network?
It is true that nations have been greatly influenced by deepening globalization, however their own borders and unique systems and ways of management in fact set the threshold for their city networks at the national scale (Fischer et al., 2006; Vinciguerra et al., 2010). This threshold will either promote or limit the internal and external flows of the whole nation. Data on global flows cannot accurately express national-scale city linkages due to threshold limitations (Bassens et al., 2010). As such, the dependency of national-scale city network studies on research into the world city network is one-sided. Research at different spatial scales is therefore necessary and crucial to address the specified city networks of disparate nations (Jacobs, 2012). Further, only a very limited number of cities have been selected as representative in existing world city network research. The Chinese city network especially cannot, however, be manifested by just a few international cities. As such, we argue that research on the Chinese city network needs to include more cities and use data from its own state.
Secondly, existing city network research has paid little attention to inter-city knowledge flows (Lu and Huang, 2012; Matthiessen et al., 2010). The world city network has been widely researched using the relational data of international advanced producer service firms (Taylor et al., 2002). However, city networks are constructed around many different flows (Pflieger and Rozenblat, 2010) - for instance, flows of capital, information, technology, knowledge and talent - of which the service flow between international firms is just one type (Zhen et al., 2012). Moreover, city networks formed by different flows can be quite distinct from one another, and the importance of different flows is subject to change within different development stages. Scholars have carried out research based on a range of different kinds of flow data, including: inter-city railway flow (Neal, 2010), aviation flow (Derudder and Witlox, 2005), port flow (Jacobs et al., 2010), internet (Vinciguerra et al., 2010) and telecommunication flow (Choi et al., 2006). Much research on Chinese city networks also has been conducted based on capital flows and material flows. A large number of studies in the past decade have also addressed knowledge flows and knowledge networks (Bornmann et al., 2011; Hennemann, 2011; Kamalski and Kirby, 2012; Martin and Moodysson, 2013). However, most of these studies consider researchers and research institutions as nodes for knowledge network; less attention has been paid to cities as nodes for knowledge networks (Liefner and Hennemann, 2011; Kratke, 2010; Ma et al., 2014; Li and Phelps, 2018).
It is valuable and meaningful to explore cities in knowledge networks (and thus “inter-city knowledge networks”). A research team organized by Matthiessen in fact addressed this theme by using the immaterial flow of “co-authorship” as the basis for their analysis (Matthiessen and Schwarz, 1999; Matthiessen et al., 2002, 2010), arguing that a strong relationship exists between city networks and scientific collaboration (co-authorship). For example, research and knowledge based on local cultures and traditions bears a strong territorial mark - in fact, these researchers maintain that the territorial base is very important for the production and distribution of knowledge (Matthiessen et al., 2002). Matthiessen et al. (2010) also point out that interaction between researchers often motivates well-established and long-lasting interaction between cities. By taking cities as nodes, the “scientific collaboration network” is able to explain the involvement of cities in knowledge-producing parts of the world economy. Within such a network, the presence of linkages between cities through flows of co-authorship indicates that a city has a high-quality value-adding capacity in the global knowledge economy system. These studies provide a better understanding of the knowledge flows and transfers that occur between cities, as well as the formation of scientific research centers.
Promoting scientific collaboration between cities is important. On the one hand, established intra-city scientific communication and knowledge spillover may make researchers within a city blind or unresponsive to external changes (Liefner and Hennemann, 2011). In such situations, inter-city scientific collaboration may reduce redundant knowledge and path lock-in (Potter and Wattsy, 2011), thereby increasing urban innovation capacity. On the other hand, encourage inter-city scientific collaboration may provide more urban innovation opportunities. Bathelt et al. (2004) highlight the necessity of knowledge inflow from distant knowledge pools and view long-distance knowledge transfer as a necessary complement to “local buzz”. An external knowledge injection contributes to the diversity of the local knowledge pool (Potter and Wattsy, 2011). As such, we believe that inter-city scientific cooperation constitutes an important new dynamic in city networks in the era of knowledge economy. This paper claims that regarding the city as a basic node for knowledge-based network research has practical applications in relation to the new goal of building an innovation-oriented nation in China. It also has the potential to provide new content and a new perspective to city network research and may enrich city network theory.
Finally, we note an apparent lack of concern in city network studies with respect to inter-city spatial interactions. With the shifts in spatial organization (from local space to flow space), our perception of the city today also needs to change - rather than the region, we therefore need to think in terms of the spatial network. A number of researchers are dealing with this shift. For example, Liefner and Hennemann (2011) proposed four types of urban regions based on actors that are located in agglomerations and occupy structural hole positions, namely: knowledge-access agglomeration, knowledge gateway, bypassed agglomeration, and peripheral utilizer. Their research demonstrated how China’s urban regions participate in national and international knowledge exchanges. Further, applying a clique analysis and a series of different thresholds, Derudder and Taylor (2005) divided several world-city cliques, identifying clique-by-clique co-membership. They found that city dyad and inter-city links at the global level retain an important regional dimension. Whilst more and more researchers are paying closer attention to the spatial characteristics of city networks, few however have considered fully the spatial interrelatedness of those networks. For example, hierarchical divisions of city networks are often based on the administrative properties or the economic strength (Zhen et al., 2012), neglecting the interactions between different nodes, which can result in a misleading view of a city’s status in the city network. In addition, the current methods used to discuss the cohesive subgroups in city networks (Derudder and Taylor, 2005), such as probability statistics and social network analysis (Yang et al., 2008; Dodds et al., 2003; Strogatz, 2001), hardly show spatial characteristics (Derudder and Taylor, 2005).
Based on the above critiques, this paper argues that city networks based on knowledge flows constitute an under-explored issue, the study of which can shed light on the transformation of city networks in globalizing urban economies. By using inter-city co-authored papers data, this paper analyzes the Chinese scientific collaboration city network, discussing its spatial features, determining factors, and points of differentiation with respect to other city networks. This paper reveals the unique characteristics of the Chinese scientific collaboration city network, perhaps providing a new perspective for city network studies.

3 Data and methodology

Knowledge production and distribution constitute a significant network. The main purpose of this paper is to elaborate the Chinese city network based on knowledge flow that constructed by scientific collaborators. Co-authored papers, as the result of scientific collaboration between researchers in different cities, present one of the most important types of knowledge flow. The data on these co-authored papers is attainable and objective. Thus, this study uses co-authored papers by researchers from different cities to represent inter-city knowledge flows (Ma et al., 2014; Li and Phelps, 2018).

3.1 Data collection

The co-authors of each paper can be regarded as a small group in those knowledge flows (Yu et al., 2011). Authors of one paper meet, discuss, work and search information online together. As a result, they produce a qualified research paper (Matthiessen et al., 2010). Working on a joint publication and sharing ideas between co-authors is usually a major form of interaction, resulting in new combinations of existing knowledge (Liefner and Hennemann, 2011). The knowledge flows between every pair or group of co-authors are, in this study, seen as knowledge flows between the two (or more) cities where the co-authors live (Kamalski and Kirby, 2012; Matthiessen et al., 2010). We made a map of the city network based on scientific collaboration, by applying a large number of those co-authored papers (knowledge flows).
Our data were obtained from the China Academic Journal Network Publishing Database (CAJD, http://www.cnki.net), the most comprehensive database of Chinese academic knowledge production in the world, which covers various fields such as the natural sciences, engineering technology, agriculture, philosophy, medicine, and the social sciences. In December 2017, the CAJD have collected more than 8000 types of academic journals, 96% of which are core journals, including more than 45 million full-text articles. Articles that were published from 2014 to 2016 and cited more than three times (on June 6, 2017) were first selected from the CAJD. Based on this data, we filtered co-authored papers and created a network matrix to represent the scientific collaboration linkages between cities.
It was still difficult to show a clear network by taking all these cities into account; as such, selecting representative cities was important. Firstly, 31 capital cities were selected, not only for considering the regional balance, but also for their importance as centers of scientific institutions and researchers. Then another 29 cities were selected based on their scientific productive ability. To make sure the representativeness of the selected 60 cities, we tested all the 334 prefecture-level cities in China to identify the main scientific centers, and affirmed that the selected 60 cities represent the top 40 or 50 largest centers in China. As such, 60 cities were finally selected as nodes to show China’s scientific collaboration network. Whilst previous studies of the Chinese city system have discussed inter-city relations under the unified system, current studies of the Chinese city network (which form a continuation of those earlier studies) no longer consider Hong Kong, Macau, or Taiwan - as such, these nodes are not addressed within this paper. Scientific collaboration linkages matrix between the 60 selected cities were created for following research.

3.2 Main indicators

(1) Inter-city Network Connectivity Degree (INCD)
The INCD between city i and city j reflects the linkage intensity of scientific collaboration between two cities in the city network.
$INC{{D}_{ij}}={{{V}_{ij}}}/{max({{V}_{ij}})}\;\times 100,\ (i,\ j=1,\ 2,\ 3,\ \ \cdots ,\ 60,\ i\ne j)$ (1)
where INCDij is the inter-city network connectivity degree between city i and city j, after standardization; Vij is the original scientific collaboration quantity obtained by accumulating all the inter-city linkages of authors worked in city i and city j; and i≠j excludes all the collaborations within a city.
(2) City Total Network Connectivity Degree (CTNCD)
The CTNCD reflects a city’s connectivity degree, indicating the overall status of the city within the network.
$CTNC{{D}_{i}}=\sum\limits_{j}{INC{{D}_{ij}},(i,\ j=1,\ 2,\ 3,\ \cdots ,\ 60,\ i\ne j)}$ (2)
where CTNCDi refers to the connectivity degree of city i, which takes one city at a time, and then aggregates all the inter-lock linkages.
(3) Clustering Coefficient
The clustering coefficient is a measure of the degree to which nodes in a graph tend to cluster together, defined as
Ci=2LN/ki(ki-1) (3)
where LN denotes the number of direct links connecting the ki nearest neighbors of city i, is equal to 1 for a city at the center of a fully inter-linked cluster, and is 0 for a link that is part of a loosely connected group.
Clustering coefficient C of a network can be defined as
C=∑k(k-1)P(K)C(K) (4)

4 Results and discussion

By using ArcGIS software, we visualized the main structure of the Chinese scientific collaboration city network. The size of the circles in Figure 1 indicates the CTNCD of the city; the width of the lines indicates the intensity of inter-city connectivity. Overall, inter-city scientific collaboration in China shows a networked pattern, with networked hierarchy, clustering features, and regional differences in the city network.
Figure 1 The Chinese city network constructed by scientific collaborators during 2014-2016

4.1 Networked hierarchy

Inspired by the global system of major scientific knowledge centers (Matthiessen et al., 2010), this paper presents a tentative categorization of Chinese cities within a system of scientific knowledge centers.
Firstly, we analyzed the distribution of 60 cities according to their CTNCD. The results show that China’s scientific collaboration city network displays a hierarchical structure. We divided the 60 nodes into six levels, in accordance with the node’s CTNCD value and the slope changes of the points (Figure 2). We then used Ucinet software to visualize the continuous attribute of 60 city nodes by taking both CTNCD and INCD into account. Similarly, we divided the 60 cities into six levels based on the distributing interval (Figure 3). In contrast to the previous division, this latter division emphasized the interactions between cities.
Figure 2 The hierarchical analysis of Chinese city network constructed by scientific collaborators during 2014-2016
Figure 3 Group map by continuous attribute based on CTNCD and INCD of Chinese city network constructed by scientific collaborators during 2014-2016
In general, both distribution methods suggested that the ranks of certain cities in the knowledge network diverged from their ranks in the administrative management hierarchy in China. In the jurisdiction-based hierarchy, the 60 cities could be divided into three levels: China’s capital (Beijing), the provincial-level cities (such as Guangzhou, Xi’an, Jinan, and Fuzhou), and the prefectural-level cities (such as Qingdao, Shenzhen, Luoyang, and Zhongshan). Not surprisingly given its very high general network connectivity, Beijing formed the core of all cities in the hierarchy of the scientific collaboration network. Most provincial-level cities were spread across the second and third levels in this scientific collaboration hierarchy, which we define as sub-national network centers and regional network centers. This finding is consistent with the fact that cities with higher ranks in the administrative levels often possess higher positions in the scientific collaboration network. However, we also note that some cities within the same administrative level had quite distinct CTNCDs, and were distributed at different levels in the scientific network (such as Jinan and Fuzhou). Importantly, this indicates that cities at the bottom of the administrative hierarchy are no longer isolated from/left behind by the surrounding cities or other higher-level cities: their increasing scientific collaborations with other cities has gradually changed their status in the city network, allowing them to play new roles in the city network of knowledge. In conclusion, unlike the claims of previous research (Ma, 2005; Wu, 2002), our analysis indicates that a city’s rank is no longer determined by its administrative level: China’s urban hierarchy is therefore shifting from a strict administrative management hierarchy to a networked (hierarchy) structure.
Furthermore, by comparing the two different distributions, we found that the ranks of certain cities in the city network changed in response to INCD (see Table 1). For example, Hefei’s rank in the city network ascended from sub-regional network center to regional network center; Shijiazhuang, Guiyang, and Hohhot’s ranks descended from sub-regional network centers to local network centers; and Quanzhou’s rank rose from local network node to local network center. In contrast, Dongguan’s rank dropped from local network center to local network node, due to its weak connectivity with other cities. We also note that results differed in accordance with the distribution method.
Table 1 Urban hierarchy distribution of Chinese city network constructed by scientific collaborators during 2014-2016
Urban Hierarchy Distribution based on CTNCD Distribution based on CTNCD and INCD
CTNCD List of cities List of cities
National network center >600 Beijing Beijing
Sub-national network center 200-600 Shanghai, Nanjing, Guangzhou, Wuhan, Hangzhou, Chengdu Shanghai, Nanjing, Hangzhou, Wuhan, Guangzhou, Chengdu
Regional network center 100-200 Xi'an, Chongqing, Tianjin, Changchun, Changsha, Lanzhou, Zhengzhou, Shenyang, Harbin Xi'an, Zhengzhou, Changsha, Harbin, Lanzhou, Hefei, Tianjin, Chongqing, Shenyang, Changchun
Sub-regional network center 50-100 Jinan, Kunming, Nanchang, Hefei, Qingdao, Dalian, Suzhou, Urumqi, Nanning, Taiyuan, Shenzhen, Xiamen, Shijiazhuang, Wuxi, Hohhot, Guiyang Qingdao, Kunming, Nanning, Taiyuan, Jinan, Wuxi, Dalian, Nanchang, Shenzhen, Xiamen, Suzhou, Urumqi
Local network center 10-50 Fuzhou, Yinchuan, Xuzhou, Yantai, Ningbo, Tangshan, Changzhou, Guilin, Xining, Daqing, Zhenjiang, Haikou, Luoyang, Qinhuangdao, Jingzhou, Lhasa, Lianyungang, Wenzhou, Baotou, Foshan, Dongguan Shijiazhuang, Yinchuan, Guiyang, Yantai, Xining, Hohhot, Xuzhou, Fuzhou, Luoyang, Changzhou, Guilin, Haikou, Lhasa, Tangshan, Ningbo, Zhenjiang, Qinhuangdao, Jingzhou, Lianyungang, Wenzhou, Daqing, Quanzhou, Foshan, Baotou
Local network node <10 Zhongshan, Quanzhou, Jiaxing, Zhuhai, Baoji, Liuzhou, Jilin Jiaxing, Zhuhai, Zhongshan, Baoji, Dongguan, Jilin, Liuzhou
Finally, what are the major factors that influence cities’ ranks in China’s scientific collaboration network? How are they affected by CTNCD? According to Lambooy (2002), knowledge is a main source of economic growth for urban regions, and in turn the economic strength of urban regions offers effective contexts for the development of knowledge. Hence, we assumed that economic strength (specifically GDP) would be an important factor in the scientific collaboration ability of a city. Population size, as a traditional criterion defining the city hierarchy, was another factor that was selected to enable comparison (Zhou and Yang, 1987). The uneven distribution of researchers and scientific resources has generated a number of attractive scientific centers, which stimulate researchers to seek collaborators there. In addition, the previous research on the regional scientific collaboration has found that regional scientific productivity could affect the collaborative preference, and the number of published articles is a determining factor (Matthiessen, 2010). Therefore, GDP (X1), population size (X2), the number of researchers (X3), and the number of published articles (X4) were chosen as factors for study. In order to discuss how they impact on a city’s scientific collaboration capacity, we undertook a multiple regression analysis between the four factors (X) and CTNCD (Y). The multiple regression equation is:
$Y=11.371+0.022{{X}_{4}}+24.887{{X}_{3}}-0.019{{X}_{1}}$ (5)
(10.623***) (3.702***) (-3.790***)
Adj R2=0.871 DW=1.683
Notes: *** indicates statistical significance at the 1% level. Adj R2 =0.871 indicates that the regression line maintains a good fit for the data. This means that the three dependent variables -X1, X3, and X4 influence the independent variable Y’s changes significantly, which can be explained at about 87.1%. DW=1.683 means that it almost excludes residual autocorrelation.
From the regression analysis, we found that: (1) population size (X2) was excluded in the regression analysis, which demonstrates that population size cannot explain the hierarchy of a city in the scientific collaboration network. As such, we believe that defining an urban system on the basis of population size is outdated. We therefore need to explore new ways to define the urban system in the era of “flow space”. (2) GDP (X1) and the number of published articles (X4) both had little impact on a city’s scientific collaboration capacity. Thus, studies that use economic strength and gravity models to simulate a city network structure could be biased (Leng et al., 2011). (3) Finally, the high correlation between X3 and CTNCD (Y) indicates that that the number of researchers plays a crucial role in contributing to a city’s scientific collaboration capacity. As such, cities should make greater efforts to attract researchers, rather than stressing the limitations of population size and weak economic strength, in order to improve their scientific collaboration capacity and rank in the city network.

4.2 Clustering regions

Another important feature of the knowledge-based Chinese city network is clustering regions. From the analysis of city nodes’ CTNCD and the probability distribution of the clustering coefficient, we found that China’s urban scientific collaboration network exhibits the feature of clustering. A scale-free network is a network whose degree distribution follows a power law (Barabasi and Albert, 1999). By calculating the probability of a city’s CTNCD, this study discovered that China’s urban scientific collaboration network is consistent with the law of a scale-free network (Figure 4). In other words, the distribution of CTNCD was found to be very uneven, with most cities having low CTNCDs, and a few having high CTNCDs. This indicates that the knowledge-based Chinese city network is neither a random network nor a perfect network, but a network with concentration centers (hub cities). Further analysis showed that the logarithms of CTNCD of the nodes in this network and the clustering coefficient maintained a negative correlation (Figure 5) - that is, some sub-networks in this scientific collaboration network were composed of highly inter-connected city nodes. This finding supports Strogatz’s discovery of the clustering feature of networks (Strogatz, 2001).
Figure 4 Probability distribution of nodal CTNCD of Chinese city network constructed by scientific collaborators during 2014-2016
Figure 5 Distribution of clustering coefficient-CTNCD correlation of Chinese city network constructed by scientific collaborators during 2014-2016
Rather than using the social network analysis method-which, whilst consistent with the preceding cluster analysis, would have been unable to adequately reflect the agglomerations spatially-we visualized the city network through a spatial distribution map. A number of the identified city clusters exhibited both relatively high linkages with outside regions and strong inter-city linkages-we term these as “regional cohesive subgroups” (Derudder and Taylor, 2005). Based on Figure 1 and previous research (Zhen et al., 2012; Lu et al., 2013), we intuitively differentiated seven city regions that demonstrated relatively strong scientific cooperation (see Table 2). By comparing these regions’ average external linkages with average internal linkages, we found that these regions have more average internal linkages (14.21) than external linkages (8.69). Identically, the average internal linkage degrees of these regions (14.43) were also higher than their external linkage degrees (10.43) (see Table 3). The presence of more and stronger internal linkages than external ones in these seven regions demonstrates the feature of clustering in the region. Meanwhile, the cities that clustered as one connected region were found to be (mostly) geographically close to each other. This confirms that geographical proximity plays a role in scientific collaboration (Katz, 1994).
Table 2 Node cities in seven regional cohesive subgroups of Chinese city network constructed by scientific collaborators during 2014-2016
Region Node cities
Beijing-Tianjin-Hebei region (BTHR) Beijing, Tianjin, Shijiazhuang, Tangshan, Qinhuangdao
Yangzi River Delta region (YRDR) Shanghai, Nanjing, Hangzhou, Suzhou, Wuxi, Ningbo, Changzhou, Zhenjiang
Pearl River Delta region (PRDR) Guangzhou, Shenzhen, Foshan, Dongguan
Chinese Central Triangle region (CCTR) Wuhan, Changsha, Nanchang, Jinzhou
Southwest region (SWR) Chengdu, Chongqing, Kunming, Guiyang
Northeast region (NER) Changchun, Shenyang, Harbin, Dalian, Daqing
Northwest region (NWR) Xi’an, Lanzhou, Yinchuan, Xining
Table 3 Comparative analysis of the seven regional cohesion subgroups of Chinese city network constructed by scientific collaborators during 2014-2016
Region Extra Region (ER) Intra-Region (IR) Gap (IR minus ER)
Line INCD Line INCD line INCD
Beijing-Tianjin-Hebei region 14.69 28.91 15.60 25.32 0.91 -3.59
Yangzi River Delta region 16.77 15.07 22.75 28.74 5.98 13.67
Pearl River Delta region 5.07 5.03 9.75 8.08 4.68 3.05
Chinese Central Triangle region 7.02 7.22 9.75 5.77 2.73 -1.45
Southwest region 6.76 6.58 16.25 16.15 9.49 9.58
Northeast region 4.94 5.33 15.60 11.21 10.66 5.88
Northwest region 5.59 4.86 9.75 5.77 4.16 0.91
Average Value 8.69 10.43 14.21 14.43 5.52 4.00
It is noted that the development degrees of the network in these seven regional cohesive subgroups appeared to be quite distinct from one another. By comparing the internal and external linkages and linkage degrees of the subgroups, we found that: (1) Whilst the Beijing-Tianjin-Hebei region (BTHR) was found to have a higher internal linkage than external, the region’s average internal INCD was weaker than its external INCD. This indicates that the scientific collaboration network in BTHR is not well developed. (2) The scientific collaboration network in the Yangzi River Delta region (YRDR) was found to be the most developed of the seven regions. YRDR was found to not only have on average higher internal linkages and INCD than external linkages and INCD, but was also the highest among the seven regions.
(3) Although both the internal linkages and INCD were higher than external linkages and INCD in the Pearl River Delta region (PRDR), their average values were lower than the average values for all of the regions. This indicates that the regional network in PRDR is not well developed. (4) Results also indicate that regional networks in the Chinese Central Triangle region and the Northwest region are not well developed. In comparison, the regional network in the Northeast and the Southwest regions appear relatively developed.
The results presented in this paper share a number of consistencies with, but also harbor a number of clear differences from, previous studies of the Chinese city network which based their analysis on various flows (Zhen et al., 2012). Like those other studies, we found the BTHR, YRDR, and PRDR to be the three most intra-connected regions. However, the status of these three regions varies in different city networks. For an example, the PRDR’s rank in the scientific collaboration city network was found to be lower than its ranks in city networks based on economic connections (Leng et al., 2011), social communications (Zhen et al., 2012), and producer services. This finding indicates that the PRDR needs to improve scientific collaboration between its cities within the region: this may become an important motivation for future regional innovation and development. By comparing city networks shaped by different flows, the strengths and weaknesses of different cities or regions can be easily uncovered, providing a better understanding of their development on a number of fronts. The differences revealed here between the city network based on knowledge flow and city networks based on other flows demonstrate that knowledge flow has become the new driving force shaping Chinese city networks.

4.3 Regional differences

China is a country with strong regional differences. Do regional differences also exist in the scientific collaboration city network? How will those differences contribute to China’s innovative ability?
We noted a series of differences when comparing the degree of connectivity of the node cities within the “traditional” division of Eastern China, Western China, and Central China (each of which are herein referred to as “regional networks”). We divided the 60 selected cities via the traditional three-part division, and then sorted the resulting three regional networks (that is, Eastern, Western, and Central China) from high to low with regard to their respective INCD (Figure 6). As a result, the INCD of the cities in the Eastern China regional network was proved to be relatively high as a whole; the average INCD of Eastern China’s 37 cities was 109.65. The INCD of the cities in Central China came next in the ranking: the regional network’s 8 node cities had an average INCD of 95.81. Finally, the INCD of the cities in Western China was proved to be relatively low; its 15 node cities had an average INCD of 71.88. Although Central China and Western China both accommodated cities with relatively high INCD, the gap between the cities of the Central China and Western China regional networks and those of Eastern China is undeniable. In addition, Beijing had the highest INCD, not only among the cities in the Eastern China regional network, but also in the whole national urban network. Undoubtedly, Beijing is the scientific center of the city network based on scientific collaboration. However, the gap between Beijing and the second city is obvious. The situation indicates that China is lacking in regional scientific centers - in other words, regional scientific centers do not perform their appropriate function.
Figure 6 City hierarchy distribution ranks in Western, Central, and Eastern China of Chinese city network constructed by scientific collaborators during 2014-2016
Differences also existed in terms of inter-region connectivity between the Western, Eastern, and Central China regional networks. As can be seen in Figure 7, Eastern China constituted the major inter-city connectivity network within the Chinese city network. The Eastern China regional network had 91 linkages, comprising 44.55% of the total linkages that make up the Chinese city network; further, the accumulated INCD in Eastern China was 1220.88, accounting for 51.38% of the total INCD of the whole network. The number of linkages between Eastern China and Central China was 28, accounting for 13.86% of the total linkages, and the accumulated INCD was 396.15, comprising 16.67% of the total INCD. The number of linkages between Eastern China and Western China was 44, accounting for 21.78% of the total linkages, and the accumulated INCD was 520.88, accounting for 21.92% of the total INCD. With respect to the direction of connections, Figure 7 also illustrates that connections from Central China and Western China toward Eastern China focused on the BTHR, the YRDR, and the PRDR, while the connections from Eastern China to Western and Central China were relatively scattered, covering all the provinces in Central and Western China. The connections between Central China and Western China were relatively weak - not only were there less linkages between them (comprising 6.93% of the total linkages), but also a lower amount of accumulated INCD (75.82, comprising 3.19% of the total INCD). These figures demonstrate that Central China has failed to play a role in connecting Eastern and Western China.
Figure 7 The network connections between Western, Central, and Eastern China of Chinese city network constructed by scientific collaborators during 2014-2016
Based on the analysis above, we suggest that China should engage in the development of regional and sub-regional scientific centers in order to achieve the goal of building an innovative country. This would enable those regional and sub-regional centers to play a bridging role between the national scientific center and local centers, which would help spread advanced knowledge and improve the benefit efficiency of scientific resources. Meanwhile, due to their restricted knowledge level, distance from, and cultural differences with the national center, the cities with low INCD could begin to cooperate with regional scientific centers to improve their scientific collaboration network.

5 Conclusions

In the process of building an innovation-oriented nation, both scientific collaboration and knowledge exchange between cities have become more frequent. Scientific collaboration between cities plays a significant role not only in alleviating the uneven distribution of scientific and technological resources but also in increasing the efficiency of those resources. This paper has provided an analysis of the spatial features of the Chinese city network based on co-authored papers data obtained from CAJD during 2014-2016; it has intuitively and objectively depicted the degree of connectivity characterizing major cities in China, and described several features of the network.
(1) We found that when based on scientific collaboration, the city network demonstrates an apparent hierarchy. However, the hierarchy based on knowledge flow diverges from the hierarchy based on traditional administrative management. In the Chinese knowledge-based city network, cities at the same administrative level could have very distinct CTNCD, and be distributed at different levels in the hierarchy. Importantly, the spatial organization of scientific cooperation amongst Chinese cities is thus shifting from a jurisdiction-based hierarchical system to a networked system. From the preceding analysis, we can see that even the cities at the bottom of the administrative hierarchy are making more connections with those cities in the higher levels, and that the small and less-developed cities also contribute more specialized functions.
(2) Highly intra-connected city regions were found to exist in the knowledge-based Chinese city network. Such regions have more average internal linkages (14.21) than external linkages (8.69), and higher average internal linkage degrees (14.43) than external linkage degrees (10.43). Whilst the knowledge-based city network appears to largely coincide with other city networks, another important finding of this study lies in the identification of differences between the knowledge-based city network and those other networks. For instance, the position of the PRDR and its cities in the knowledge-based urban network was shown to be relatively lower in our study than in other studies based on other types of urban network. We therefore suggest that policy makers in the PRDR should place greater focus on inter-city scientific collaborations, which could promote the city’s innovative capacity. Through research into knowledge flows, we can gain a more comprehensive analysis of the formation mechanism of urban networks.
(3) The previous Chinese city network studies suggest that a city’s population size and economic strength are the two most important factors in deciding a city’s rank in the network. However, the results of the regression analysis detailed here demonstrate that a city’s population size and economic strength in fact have little impact on a city’s rank in the city network of knowledge. The number of researchers was rather found to be the crucial factor in a city’s connectivity in a network based on knowledge flow. This indicates that city networks based on different flows could be very distinct from each other, even in terms of their determining factors. Therefore, we argue that studies of city networks shaped by different flows are necessary and crucial to present the multifaceted nature of a city network. Further, we advocate that comparisons of various city networks should be undertaken. Such comparisons would be beneficial to the construction of regional networks and provide practical policy suggestions to enhance a city’s status in the network.
This paper constitutes a preliminary attempt to explore the knowledge-based Chinese city network using data on scientific collaboration. A number of important related discussions remain to be explored through future research. For instance, how can the national-scale city network link with the world city network in a better way? What are the differences between city networks on the basis of two different types of knowledge flow - for instance, technical collaboration and scientific collaboration? And, finally, can better quantitative methods be formulated in order to define and classify regional cohesive subgroups?

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
Barabasi A L, Albert R, 1999. Emergence of scaling in random networks.Science, 286(5439): 509-512.

DOI

[2]
Bassens D, Derudder B, Taylor P Jet al., 2010. World city network integration in the Eurasian Realm.Eurasian Geography and Economics, 51(3): 385-401.A team of authors from the Globalization and World Cities (GaWC) research network explores the main empirical features of the global economic integration of Eurasia through an analysis of the (shifting) position of that macroregion's key cities. More specifically, they examine the gateway roles played by key cities in four major subregions—China, the former Eastern Bloc (FEB, i.e., Central and Eastern Europe), former Soviet Union (FSU), and India—in the channeling of transnational flows of capital, goods, knowledge, and people in 2000-2008, focusing on the globalization of firms in the financial and business service sectors. Although cities in the four subregions share a common recent history of transition from closed and state-centered to open and market-based economies, the study reveals the persistence of distinctive geographical idiosyncrasies in their connections across the globe. Journal of Economic Literature, Classification Numbers: F020, F200, O180, P300. 4 figures, 3 tables, 1 appendix, 39 references.

DOI

[3]
Bathelt H, Malmberg A, Maskell P, 2004. Clusters and knowledge: Local buzz, global pipelines and the process of knowledge creation.Progress in Human Geography, 28(1): 31-56.

DOI

[4]
Bornmann L, Leydesdorff L, Walch-Solimena Cet al., 2011. Mapping excellence in the geography of science: An approach based on Scopus data.Journal of Informetrics, 5(4): 537-546.As research becomes an ever more globalized activity, there is growing interest in national and international comparisons of standards and quality in different countries and regions. A sign for this trend is the increasing interest in rankings of universities according to their research performance, both inside but also outside the scientific environment. New methods presented in this paper, enable us to map centers of excellence around the world using programs that are freely available. Based on Scopus data, field-specific excellence can be identified and agglomerated in regions and cities where recently highly cited papers were published. Differences in performance rates can be visualized on the map using colours and sizes of the marks.Graphical abstractLocations of authors in Europe having published highly cited physics and astronomy papers in 2007 (Scopus data).

DOI

[5]
Bourne L S, 1975. Urban Systems:Strategies for Regulation. Oxford: Oxford University Press.

[6]
Camagni R P, Capello R, 2004. The city network paradigm: Theory and empirical evidence. In: Capello R, Nijkamp P. eds. Urban Dynamics and Growth. Netherlands: Elsevier B.V. R.Real city systems in advanced countries have significantly departed from the abstract Christallerian pattern of a nested hierarchy of centers and markets. The reduction in transport costs and the demand for ‘variety’ by the consumer have broken the theoretical hypothesis of separated, gravity-type, non-overlapping market areas. ‘Location economies’ as described by Hoover, and synergy elements operating through horizontal and vertical linkages among firms, have led to the emergence of specialized centers, in contrast to the typical despecialisation pattern deriving from the theoretical model. High-order functions sometimes locate in small (but specialized) centers where the model expectations are only for lower-order functions. In recent years, network behaviour has extensively been analysed as the emerging model for economic growth. By network behaviour, a metaphor for cooperative behaviour among individuals, corporate or territorial partners is intended - which is increasingly becoming the reference paradigm in an era of continuing innovation and fast technological change, in the presence of “market failure” when dynamic and innovative behaviours are concerned and of the high costs of a growth strategy based on the sole internal know-how. The chapter provides insights into the economic rational of this behaviour, by approaching the issue treating cities and territories both as individuals and as collective actors. Moreover, the chapter provides empirical evidence on the existence of “city networks” and on the advantages cities gain from being part of a network.

DOI

[7]
Castells M, 1989. The Information City. Oxford:Basil Blackwell.

[8]
Choi J H, Barnett G A, Chon B S, 2006. Comparing world city networks: A network analysis of Internet backbone and air transport intercity linkages.Global Networks: A Journal of Transnational Affairs, 6(1): 81-99.Abstract The research looks at the structure of the Internet backbone and air transport networks between 82 cities in 2002, using Internet backbone bandwidth and air passenger traffic data. Centrality measures on individual city's hierarchy in the Internet and in the air traffic networks were significantly correlated, with London in the most dominant position in both networks. A quadratic assignment procedure (QAP) showed a structural equivalence between two systems. The division and membership of the clusters in both networks also showed similarity; both networks had a strongly cohesive North American-European cluster with the London-New York dyad as the strongest linkage in the global flow of information and people. These findings suggest that current trends in Internet infrastructure concentration reproduce and maintain global inequality and hierarchy among world cities.

DOI

[9]
Derudder B, Taylor P J, 2005. The cliquishness of world cities.Global Networks, 5(1): 71-91.Abstract Through an exploratory analysis of the world city network, we confront the dearth of studies on inter-urban networks on a global scale. In the first section, we outline how a formal theoretical specification of the world city network as a social network has guided an extensive data gathering effort. Through this specification it becomes possible to apply standard techniques of network analysis to world cities. In the second section we outline a framework within which to explore the complex patterns that arise from the formal specification of the world city network. The conjectural use of connectivity, world city cliques and clique-by-clique comembership is then used in the third section as the input to an exploratory analysis of the measures of connectivity at different thresholds. The main findings are the two city dominance of the New York London dyad, and the observation that inter-city links at the global level retain an important regional dimension. Exemplary cases are the thorough interrelatedness of North American, Pacific Asian and European world cities.

DOI

[10]
Derudder B, Taylor P J, Hoyler Met al., 2013. Measurement and interpretation of connectivity of Chinese cities in world city network, 2010.Chinese Geographical Science, 23(3): 261-273.

DOI

[11]
Derudder B, Witlox F, 2005. An appraisal of the use of airline data in assessing the world city network: A research note on data.Urban Studies, 42(13): 2371-2388.

DOI

[12]
Dodds P S, Muhamad R, Watts D J, 2003. An experimental study of search in global social networks.Science, 301(5634): 827-829.We report on a global social-search experiment in which more than 60,000 e-mail users attempted to reach one of 18 target persons in 13 countries by forwarding messages to acquaintances. We find that successful social search is conducted primarily through intermediate to weak strength ties, does not require highly connected "hubs" to succeed, and, in contrast to unsuccessful social search, disproportionately relies on professional relationships. By accounting for the attrition of message chains, we estimate that social searches can reach their targets in a median of five to seven steps, depending on the separation of source and target, although small variations in chain lengths and participation rates generate large differences in target reachability. We conclude that although global social networks are, in principle, searchable, actual success depends sensitively on individual incentives.

DOI PMID

[13]
Esparza A X, Krmenec A J, 2000. Large city interaction in the US urban system.Urban Studies, 37(4): 691-709.Mounting interest in the global system of cities has led many scholars to portray large, world-class cities as control points in the global economy. From this position, they govern global banking, finance and international corporate relations with the assistance of advanced business services and telecommunications. While large cities have taken on many global functions, recent research has cast them only as outposts in the global economy and pays little attention to their role in regional and national economies. This paper explores the role of five large US cities in regional and national economies. Using primary trade data for five high-order business services, we identify the extent to which these large cities trade with other large cities of the US urban system, smaller-sized cities, and test statistically the effect that firm attributes and physical characteristics of the city system have on shaping the areal extent of large city interaction (trade). The results indicate that: trade is consistently biased towards other large cities; the size of origin cities is associated with the size of destination cities; and, distance more so than destination city size plays the dominant role in shaping large city patterns of trade.

DOI

[14]
Fang C, Ma H, Wang Z B, Li G D, 2014. The sustainable development of innovative cities in China: Comprehensive assessment and future configuration. Journal of Geographical Sciences, 24(6): 1095-1114.中国科学院机构知识库(CAS IR GRID)以发展机构知识能力和知识管理能力为目标，快速实现对本机构知识资产的收集、长期保存、合理传播利用，积极建设对知识内容进行捕获、转化、传播、利用和审计的能力，逐步建设包括知识内容分析、关系分析和能力审计在内的知识服务能力，开展综合知识管理。

DOI

[15]
Fischer M M, Scherngell T, Jansenberger E, 2006. The geography of knowledge spillovers between high-technology firms in Europe: Evidence from a spatial interaction modeling perspective.Geographical Analysis, 38(3): 288-309.The focus in this article is on knowledge spillovers between high-technology firms in Europe, as captured by patent citations. The European coverage is given by patent applications at the European Patent Office that are assigned to high-technology firms located in the EU-25 member states (except Cyprus and Malta), the two accession countries Bulgaria and Romania, and Norway and Switzerland. By following the paper trail left by citations between these high-technology patents we adopt a Poisson spatial interaction modeling perspective to identify and measure spatial separation effects to interregional knowledge spillovers. In doing so we control for technological proximity between the regions, as geographical distance could be just proxying for technological proximity. The study produces prima facie evidence that geography matters. First, geographical distance has a significant impact on knowledge spillovers, and this effect is substantial. Second, national border effects are important and dominate geographical distance effects. Knowledge flows within European countries more easily than across. Not only geography, but also technological proximity matters. Interregional knowledge flows are industry specific and occur most often between regions located close to each other in technological space.

DOI

[16]
Friedmann J, 1995. Where we stand: A decade of world city research. In: Knox P L, Taylor P J. World Cities in a World System. New York: Cambridge University Press, 21-47.

[17]
Harris C D, Ullman E L, 1945. The nature of cities.The Annals of the American Academy of Political and Social Science, 242(1): 7-17.

DOI

[18]
Hou H, Kretschmer H, Liu Z, 2008. The structure of scientific collaboration networks in Scientometrics.Scientometrics, 75(2): 189-202.The structure of scientific collaboration networks in scientometrics is investigated at the level of individuals by using bibliographic data of all papers published in the international journal Scientometrics retrieved from the Science Citation Index (SCI) of the years 1978–2004. Combined analysis of social network analysis (SNA), co-occurrence analysis, cluster analysis and frequency analysis of words is explored to reveal: (1) The microstructure of the collaboration network on scientists’ aspects of scientometrics; (2) The major collaborative fields of the whole network and of different collaborative sub-networks; (3) The collaborative center of the collaboration network in scientometrics.

DOI

[19]
Jacobs A J, 2012. The World’s Cities: Contrasting Regional, National, and Global Perspectives. London: Routledge.

[20]
Jacobs W, Ducruet C, De Langen P, 2010. Integrating world cities into production networks: The case of port cities.Global Networks: A Journal of Transnational Affairs, 10(1): 92-113.In this article we analyse the location patterns of firms that provide specialized advanced producer services (APS) to international commodity chains that move through seaports. Such activities can take place in world cities or in port cities. The analysis of APS location patterns in port cities provides a good opportunity to integrate the study of world cities into the framework of Global Production Networks. Based upon our empirical findings, we conclude that while port-related APS activities predominantly follow the world city hierarchy, a number of port cities stand out because they act as nodes in global commodity flows and as centres of advanced services related to shipping and port activities. Based on these empirical findings we address future avenues of research.

DOI

[21]
Kamalski J, Kirby A, 2012. Bibliometrics and urban knowledge transfer.Cities, 29(16): S3-S8.This paper demonstrates the potential of bibliometric analysis in the context of urban studies. After a brief discussion of the measurement of knowledge production, we provide an analysis of how the field of urban studies is constructed. We do this in three contexts: first, in the narrowly defined population of journals that constitutes the Thomson Reuters classification of urban studies; second, in the larger population of journals deemed to be within the social and behavioral sciences; and third, in a subset of the applied sciences. We find that, by using keyword analysis, it is possible to identify three distinct spheres of ‘urban knowledge’ that contain some overlap but also significant differences. We explore the significance of that for the development of urban studies.

DOI

[22]
Katz J S, 1994. Geographical proximity and scientific collaboration.Scientometrics, 31(1): 31-43.

DOI

[23]
Kratke S, 2010. Regional knowledge networks a network analysis approach to the interlinking of knowledge resources.European Urban and Regional Studies, 17(1): 83-97.

DOI

[24]
Lei X P, Zhao Z Y, Zhang Xet al., 2013. Technological collaboration patterns in solar cell industry based on patent inventors and assignees analysis.Scientometrics, 96(2): 427-441.

DOI

[25]
Leng B, Yang Y, Li Y, Zhao S, 2011. Spatial characteristics and complex analysis: A perspective from basic activities of urban networks in China.Acta Geographica Sinica, 66(2): 199-211. (in Chinese)

[26]
Leydesdorff L, Persson O, 2010. Mapping the geography of science: Distribution patterns and networks of relations among cities and institutes.Journal of the American Society for Information Science and Technology, 61(8): 1622-1634.Using Google Earth, Google Maps, and/or network visualization programs such as Pajek, one can overlay the network of relations among addresses in scientific publications onto the geographic map. The authors discuss the pros and cons of various options, and provide software (freeware) for bridging existing gaps between the Science Citation Indices (Thomson Reuters) and Scopus (Elsevier), on the one hand, and these various visualization tools on the other. At the level of city names, the global map can be drawn reliably on the basis of the available address information. At the level of the names of organizations and institutes, there are problems of unification both in the ISI databases and with Scopus. Pajek enables a combination of visualization and statistical analysis, whereas the Google Maps and its derivatives provide superior tools on the Internet.

DOI

[27]
Li Y C, Phelps N, 2018. Megalopolis unbound: Knowledge collaboration and functional polycentricity within and beyond the Yangtze River Delta Region in China, 2014.Urban Studies, 53(2): 1035-1047.

[28]
Liang L M, Zhu L, 2002. Major factors affecting China’s inter-regional research collaboration: Regional scientific productivity and geographical proximity.Scientometrics, 55(2): 287-316.On the basis of the measured frequency distribution of China"s inter-regional co-authored papers covered by the Chinese Science Citation Database , this paper shows the pattern of China"s inter-regional research collaboration (IRRC), and analyzes how the collaborative pattern was formed. A new method is used to calculate the expected value matrix based on an observed value matrix of IRRC, which is asymmetric and has no diagonal elements. The results fall into three groups. 1) Regional scientific productivity affects both the collaborative preference and ranking of authors" name; 2) geographical proximity is an important factor determining the pattern of IRRC; 3) when using Salton"s measure, regional mean collaborative strength increases as the regional productivity increases, and as the distance between two regions decreases.

DOI

[29]
Liefner I, Hennemann S, 2011. Structural holes and new dimensions of distance: The spatial configuration of the scientific knowledge network of China’s optical technology sector.Environment and Planning A, 43(4): 810-829.

DOI

[30]
Lu L C, Huang R, 2012. Urban hierarchy of innovation capability and inter-city linkages of knowledge in post-reform China.Chinese Geographical Science, 22(5): 602-616.AbstractThe 2000s has witnessed increasing interests in cities role of innovation in the era of knowledge-based economy. Compared with substantial empirical analysis on the world city hierarchy of innovation, this paper attempts to examine the national urban hierarchy of innovation capability in China, in terms of ranking systems, spatial pattern and inter-city linkages of knowledge during the post-reform period since the late 1970s. Based on quantitative analysis such as principal component factor analysis and clustering analysis, this paper identifies the five-tier hierarchy of innovation, which is headed by Beijing and Shanghai, followed by the capital cities of each province and regional centre cities. The development of China urban hierarchy of innovation capability has been driven by such factors as the scale of innovation, scientific scale, innovation potential and innovation environment. The paper further investigates the inter-city linkages of knowledge measured by the number of co-authored papers among the cities. Beijing is positioned in the central position of the knowledge diffusion and knowledge cooperation innovation. More knowledge diffusion among high level cities has occurred than that among the low level cities as well as between the low level cities and high level cities, and provincial capital cities and the regional central cities.

DOI

[31]
Lu R, Zhang R K, Reve T, 2013. Relations among clusters in six Chinese city regions.European Planning Studies, 21(8): 1189-1209.Despite the considerable number of papers that have discussed industrial clusters, it is surprising that there is little research evidence on relations among clusters. This article collects longitudinal data on three low-tech and two high-tech industrial clusters in six cities in the dynamic Pearl River Delta of the People's Republic of China. The findings provide empirical support to both the Marshall–Arrow–Romer model, which argues for the importance of homogeneous (specialized) clusters and the Jacobs model, which argues for the importance of heterogeneous (diversified) clusters. Both homogeneous and heterogeneous clusters in the same region influence one another's cluster size and economic output.

DOI

[32]
Ma H T, Fang C L, Pang Bet al., 2014. The effect of geographical proximity on scientific cooperation among Chinese cities from 1990 to 2010.Plos One, 9(11): e111705.The relations between geographical proximity and spatial distance constitute a popular topic of concern. Thus, how geographical proximity affects scientific cooperation, and whether geographically proximate scientific cooperation activities in fact exhibit geographic scale features should be investigated.Selected statistics from the ISI database on cooperatively authored papers, the authors of which resided in 60 typical cites in China, and which were published in the years 1990, 1995, 2000, 2005, and 2010, were used to establish matrices of geographic distance and cooperation levels between cities. By constructing a distance-cooperation model, the degree of scientific cooperation based on spatial distance was calculated. The relationship between geographical proximity and scientific cooperation, as well as changes in that relationship, was explored using the fitting function.(1) Instead of declining, the role of geographical proximity in inter-city scientific cooperation has increased gradually but significantly with the popularization of telecommunication technologies; (2) the relationship between geographical proximity and scientific cooperation has not followed a perfect declining curve, and at certain spatial scales, the distance-decay regularity does not work; (3) the Chinese scientific cooperation network gathers around different regional center cities, showing a trend towards a regional network; within this cooperation network the amount of inter-city cooperation occurring at close range increased greatly.The relationship between inter-city geographical distance and scientific cooperation has been enhanced and strengthened over time.

DOI PMID

[33]
Ma L J C, 2005. Urban administrative restructuring, changing scale relations and local economic development in China.Political Geography, 24(4): 477-497.China has significantly restructured its urban administrative/spatial system since the 1978 economic reforms to empower central cities to play a leading role in driving national and regional economic development. Through changes in the scale relations that are heavily conditioned by the administrative levels or ranks ( dengji) of cities and counties, central cities have in fact been transformed into city-regions. In the reorganization of urban administrative space, many counties have acquired city status while others have been turned into city districts. Such changes have clearly benefited the central cities. Key to the understanding of the tensions among the spatial units constituting the urban administrative/spatial system is the administrative levels of the units in the system which can be seen as a special type of scale. Though restructured, China's space economy continues to be governed administratively.

DOI

[34]
Martin R, Moodysson J, 2013. Comparing knowledge bases: On the geography and organization of knowledge sourcing in the regional innovation system of Scania, Sweden.European Urban and Regional Studies, 20(2): 170-187.

DOI

[35]
Matthiessen C W, Schwarz A W, 1999. Scientific centers in Europe: An analysis of research strength and patterns of specialisation based on bibliometric indicators.Urban Studies, 36(3): 453-477.This paper presents the first analysis of scientific strength by output (papers in the Science Citation Index 1994-96) produced by authors from the 'greater' urban regions of Europe. Top lists of European centres are indicated. Four agglomerations constitute the European super-league of science: London, Paris, Moscow and the Dutch urban agglomeration of Amsterdam, the Hague, Rotterdam and Utrecht. The next layer could be named the primary league and comprises 19 large research centres. A third group of 16 cities forms a secondary league of 16 smaller research centres. These upper-level research strongholds are categorised in the paper where patterns of specialisations by absolute and relative distribution of research disciplines for each city are identified and families of cities by research pattern are analysed and compared within the perspective of urban economic growth and change.

DOI

[36]
Matthiessen C W, Schwarz A W, Find S, 2002. The top-level global research system, 1997-99: Centers, networks and nodality. An analysis based on bibliometric indicators.Urban Studies, 39(1): 903-927.

DOI

[37]
Matthiessen C W, Schwarz A W, Find S, 2010. World cities of scientific knowledge: Systems, networks and potential dynamics. An analysis based on bibliometric indicators.Urban Studies, 47(9): 1879-1897.BT - Multilayered Cities and Urban Systems

DOI

[38]
Mo H H, Wang J E, Jin F J, 2009. Identifying centrality in the air transportation network of China.Transportation and Geography, 2(2): 493-502.

[39]
Neal Z, 2010. Refining the air traffic approach to city networks.Urban Studies, 47(10): 2195-2215.

DOI

[40]
Neal Z, 2011. From central places to network bases: A transition in the U.S. urban hierarchy, 1900-2000.City & Community, 10(1): 49-75.

[41]
Neal Z, 2013. Brute force and sorting processes: Two perspectives on world city network formation.Urban Studies, 50(6): 1277-1291.

DOI

[42]
Pflieger G, Rozenblat C, 2010. Introduction. Urban networks and network theory: The city as the connector of multiple networks.Urban Studies, 47(13): 2723-2735.

DOI

[43]
Potter A, Wattsy H D, 2011. Evolutionary agglomeration theory: Increasing returns, diminishing returns, and the industry life cycle.Journal of Economic Geography, 11(3): 417-455.According to Marshall?s agglomeration theory, Krugman?s New Economic Geography models, and Porter?s cluster policies, firms should receive increasing returns from a trinity of agglomeration economies: a local pool of skilled labour, local supplier linkages, and local knowledge spillovers. Recent evolutionary theories suggest that whether agglomeration economies generate increasing returns or diminishing returns depends on time, and especially the evolution of the industry life cycle. At the start of the 21st century, we re-examine Marshall?s trinity of agglomeration economies in the city-region where he discovered them. The econometric results from our multivariate regression models are the polar opposite of Marshall?s. During the later stages of the industry life cycle, Marshall?s agglomeration economies decrease the economic performance of firms and create widespread diminishing returns for the economic development of the city-region, which has evolved to become one of the poorest city-regions in Europe.

DOI

[44]
Smith D A, Timberlake M, 1995. Conceptualising and mapping the structure of the world system’s city system.Urban Studies, 32(2): 287-302.

DOI

[45]
Strogatz S H, 2001. Exploring complex networks.Nature, 410(6825): 268-276.

DOI

[46]
Taylor P J, 2004. World Cities Network: A Global Urban Analysis. London: Routledge.

[47]
Taylor P J, 2005. Leading world cities: Empirical evaluations of urban nodes in multiple networks.Urban Studies, 42(9): 1593-1608.

DOI

[48]
Taylor P J, 2012. Comment on looking back to look forward: Where are China’s cities?Eurasian Geography and Economics, 53(3): 327-330.react-text: 151 New Zealand historiography focusing on tourism has tended to favour the rural over the urban. This paper argues that cities and towns were also the focus of much publicity for promoting migration and tourism to the country. As evidence it examines a generally overlooked historical source, booster texts. It suggests that such sources should be considered a genre of travel writing on the basis... /react-text react-text: 152 /react-text [Show full abstract]

DOI

[49]
Taylor P J, Catalano G, Walker D R F, 2002. Measurement of the world city network.Urban Studies, 39(13): 2367-2376.

DOI

[50]
Taylor P J, Derudder B, 2017. World city network.Urban Geography, 38(1): 159-161.

DOI

[51]
Taylor P J, Hoyler M, Verbruggen R, 2010. External urban relational process: Introducing central flow theory to complement central place theory.Urban Studies, 47(13): 2803-2818.

DOI

[52]
Vinciguerra S, Frenken K, Valente M, 2010. The geography of internet infrastructure: An evolutionary simulation approach based on preferential attachment. Papers in Evolutionary Economic Geography, 47(9): 1969-1984.

[53]
Wang C J, Ducruet C, Wang W et al.Wang W , 2015. Evolution, accessibility and dynamics of road networks in China from 1600 BC to 1900 AD.Journal of Geographical Sciences, 25(4): 451-484.Before the emergence of modern modes of transport, the traditional road infrastructure was the major historical means of carrying out nationwide socio-economic exchange. However, the history of transport infrastructure has received little attention from re-searchers. Given this background, the work reported here examined the long-term development of transport networks in China. The national road network was selected for study and the 3500 years from 1600 BC to 1900 AD was chosen as the study period. Indicators were designed for the maturity level of road networks and an accessibility model was developed for the paths of the shortest distance. The evolution of the road network in China since the Shang Dynasty (1600 BC) was described and its major features were summarized to reveal long-term regularities. The maturity level of the road network and its accessibility was assessed and regions with good and poor networks were identified. The relationship between China's natural, social, and economic systems and the road network were discussed. Our analysis shows that the road network in China has a number of long-term regularities. The continuously expanding road network follows a path of inland expansion especially towards the border areas. However, its coverage and accessibility are characterized by a core-peripheral configuration, which has close relationships with, not only the natural conditions, but also national defense and warfare. The centralization of national power, national land governance, postal transport, the transport of specialized cargos, and international trade are also related to the development of the road network. This research draws attention to the evolving regularities of transport networks.

DOI

[54]
Wu F L, 2002. China’s changing urban governance in the transition towards a more market-oriented economy.Urban Studies, 39(7): 1071-1093.

DOI

[55]
Yang K Q, Yang L, Gong B H et al.Gong B H , 2008. Geographical networks: Geographical effects on network properties.Frontiers of Physics in China, 3(1): 105-111.Complex networks describe a wide range of systems in nature and society. Since most real systems exist in certain physical space and the distance between the nodes has influence on the connections, it is helpful to study geographical complex networks and to investigate how the geographical constrains on the connections affect the network properties. In this paper, we briefly review our recent progress on geographical complex networks with respect of statistics, modelling, robustness, and synchronizability. It has been shown that the geographical constrains tend to make the network less robust and less synchronizable. Synchronization on random networks and clustered networks is also studied.

DOI

[56]
Yu Q, Shao H F, Duan Z G, 2011. Research groups of oncology co-authorship network in China.Scientometrics, 89(2): 553-567.This paper aims at analyzing and extracting the research groups from the co-authorship network of oncology in China. By use of centrality, component analysis, K-Core, M-Slice, Hierarchical Clustering analysis, and Multidimensional Scaling analysis, we studied the data from 10 Core Chinese Oncology journals between 2000 and 2009, analyzed the structure character of the Chinese Oncology research institutes. This study advances the methods for selecting the most prolific research groups and individuals in Chinese Oncology research community, and provides basis for more productive cooperation in the future. This study also provides scientific evidences and suggestions for policymakers to establish a more efficient system for managing and financing Chinese Oncology research in the future.

DOI

[57]
Zhen F, Wang B, Chen Y, 2012. China’s city network characteristics based on social network space: An empirical analysis of Sina Micro-blog.Acta Geographica Sinica, 67(8): 1031-1043. (in Chinese)

[58]
Zhao M X, Wu K, Liu X J et al.Liu X J , 2015. A novel method for approximating intercity networks: An empirical comparison for validating the city networks in two Chinese city-regions.Journal of Geographical Sciences, 25(3): 337-354.A network perspective has increasingly become an organizational paradigm for understanding regional spatial structures. Based on a critical overview of existing empirical models for estimating intercity networks based on firm linkages, this study extends the recently proposed regional corporate city model algorithm by proposing a new method for approximating urban networks based on the locational strategies of firms. The new method considers both regional and hierarchical network features and avoids the information loss associated with the conversion from two-mode firm-city networks to one-mode city-city networks. In addition, networks estimated by using the method proposed herein are suitable when employing social network analysis. Finally, this method is empirically validated by examining intercity firm networks formed by advanced producer services firms in China two largest metropolitan areas, namely the Yangtze River Delta and Pearl River Delta. The presented empirical analysis suggests two main findings. First, in contrast to conventional methods (e.g., the interlocking city network model), our new method produces regional and hierarchical urban networks that more closely resemble reality. Second, the new method allows us to use social network analysis to assess betweenness and closeness centralities. However, regardless of the model applied, the validity of any method that measures urban networks depends on the soundness of its underlying assumptions about how network actors (firms, in our case) interact.

DOI

[59]
Zhou Y X, Yang Q, 1987. Changes in the urban hierarchy in China and the territorial types by province.Chinese Sociology and Anthropology, 19(3/4): 137-159.The urban hierarchy is an important part of a country's urban system. It reflects the distribution and law of development of urbanness in a country or a region. In determining urban population sizes and development strategies, we can avoid errors of judgment by viewing them against the overall framework of the urban system in the nation.

DOI

Options
Outlines

/

〈 〉