Public Space Optimization Strategy Through Social Network Analysis in Shenzhen’s Gongming Ancient Fair

Abstract

Ancient fairs in China were designated as commercial zones with fixed stalls that had emerged from commodity exchange demands and socio-cultural factors such as clan systems and gentry intervention, exhibiting dual commercial–communal characteristics. Several ancient fairs in Shenzhen have been retained, including Gongming Ancient Fair, which maintains its original spatial configuration adjacent to industrial zones and urban villages, attracting a high concentration of migrant workers. Survey results show that 85% of Gongming residents demand public space renovations, citing inadequacy of the spaces to support public activities. Given the intrinsic link between public spaces and public activities, fostering their positive interaction is crucial for enhancing urban vitality. However, existing studies predominantly focus on the physical environment and neglect activity-driven optimization perspectives. This study first employed social network analysis (SNA) to construct two networks of Gongming Ancient Fair, using the software Ucinet 6.755, including a public space network based on spatial connectivity and a public activity network based on pedestrian flow. Second, the networks’ structural characteristics were analyzed using seven metrics: node degree, density, betweenness centrality, betweenness centralization, clustering coefficient, average path length, and small-world property. Discrepancies between the networks were quantified through betweenness centrality comparisons, with field surveys and interviews identifying causal factors including seasonal product limitations, spatial constraints, inadequate supporting facilities, and substandard management. Based on the survey data and analytical results, the key renovation nodes were categorized into three types: high-control-capacity nodes, high-expectation nodes, and culturally distinctive nodes. Finally, three optimization strategies are proposed. This study integrates sociological perspectives into ancient fair revitalization, addressing gaps in activity-driven spatial research.

1. Introduction

In the current era of urban stock renewal, the revitalization of spatial structures in ancient fairs, as historically significant areas, holds substantial importance. Urban regeneration, particularly in historic districts or urban central areas, presents a dual challenge in balancing heritage conservation with the need for adaptive reuse [1]. Shenzhen, a primary origin of Lingnan’s ancient fairs, exemplifies the tension between urbanization and heritage preservation, with its ancient fairs presenting representative cases for research. The Shenzhen municipal government introduced policy frameworks to promote the revitalization of ancient fairs. For instance, Longhua District proposed the “Hundreds, Thousands, Tens of Thousands (Bai Qian Wan) Project” to drive development through cultural tourism. In addition, scholars have conducted extensive studies on ancient fairs in Shenzhen, exploring formation mechanisms [2,3] and renewal strategies [4,5,6,7,8,9].

Ancient fairs (or “Xushi”) refer to designated commercial zones in ancient Chinese cities. The emergence and evolution of ancient fairs resulted from the interplay between commodity exchange demands and human factors, such as clan systems and gentry intervention, exhibiting dual commercial–communal characteristics. Ancient fairs in Shenzhen exhibit diverse typologies and varying scales. Existing research [6] categorizes them into three types based on renewal levels: renovated, restored, and original-preservation ancient fairs. Renovated and restored ancient fairs have transitioned to cultural tourism industries, with spatially optimized environments oriented toward external populations such as tourists. In contrast, original-preservation ancient fairs are smaller, have fewer historical elements, and lack the short-term potential for cultural tourism transformation, retaining their original configurations. These spaces primarily serve internal residents, including indigenous inhabitants, migrant workers, and nearby laborers.

Despite policy incentives for cultural tourism transformation, most original-preservation ancient fairs remain in a transitional stagnation phase, and existing public spaces fail to accommodate the diverse activity demands of massive transient populations. While existing studies propose optimization strategies from multiple perspectives, they predominantly focus on the formation mechanisms and spatial renewal strategies of ancient fairs [2,3,4,5,6,7,8,9] and neglect the socio-spatial perspectives of public space utilization.

Social network analysis (SNA) is an emerging analytical tool in spatial studies that has been applied to historical districts through different network comparisons [10,11,12,13,14,15,16]; it demonstrates versatility under diverse research scales, offering tools to visualize the myriad relationships, organizations, and factions [17], with particular strength in analyzing the compatibility of networks. This study focuses on original-preservation ancient fairs and employs SNA to construct a public space–activity network for Gongming Ancient Fair as a case study. By comparing two network characteristics and interdependencies, this study identifies existing problems and proposes spatial optimization strategies, providing insights into the social dimensions of spatial optimization for similar historic districts undergoing industrial transformation in other cities.

2. Literature Review

Existing studies primarily apply SNA to examine social relations among stakeholders in historic districts [18,19], while its application to spatial research remains limited. Analyzing the relevant literature could offer insights into SNA applications in spatial optimization in international historic districts.

2.1. Public Space Optimization Strategy in the Ancient Fair

The ancient fair is a designated area for commodity trading in ancient Chinese cities and typically features fixed stalls or shops. As a historical space with significant value, the renewal of public spaces in ancient fairs has attracted increasing attention.

Regarding formation mechanisms, Wang J. et al. [2] developed a dialectical framework of “power–capital–society” based on the spatial production theory of Lefebvre, revealing the dynamic mechanisms of Shajing Ancient Fair’s renewal and providing guidance for ancient fairs’ public space revitalization. Ding X. [3] employed the stratigraphic approach of urban historical landscapes to demonstrate the collaborative dynamic formation mechanisms involving natural foundations, economic patterns, and clan culture in Shajing Ancient Fair, proposing an “anchor-stratification” model and functional activation strategies.

In terms of renewal strategies, Tong D. et al. [4] established a research framework for revitalizing historical districts through spatial, cultural, industrial, and social dimensions based on holistic conservation theory, using Shenzhen Qingping Ancient Fair as a case study to analyze the connotation, characteristics, and practical applications of integrated conservation. Huang J. W. et al. [5] developed a landscape feature evaluation system across natural-culture, artificial-construction, spatial-organization, and environmental-landscape dimensions in Qiao Fair, Taishan, proposing differentiated enhancement strategies. Li J. [6] established a tripartite attribute framework involving commercial, communal, and cultural characteristics for fair spaces, deriving renewal strategies through empirical case studies at urban block, architectural, public space, and intangible cultural levels.

Some studies focus specifically on spatial design strategies. Qiu H. K. et al. [7] emphasized boundary space design in ancient fairs, proposing renewal strategies centered on “porosity, permeability, and locality”. Zhang T. T. et al. [8] analyzed fire safety principles through case studies, developing evacuation design strategies by using Shenzhen Gongming Ancient Fair as an example. Zhang Y. X. et al. [9] implemented micro-renewal interventions in Shajing Ancient Fair based on daily life authenticity and spatial value preservation.

While existing research proposes multi-dimensional optimization strategies for physical spaces, studies addressing spatial improvements based on residents’ public activities remain insufficient. Given the crucial interaction between public spaces and communal activities for urban vitality enhancement and spatial optimization [20], this study focuses on the public activities of ancient fair residents, proposing existing issues and corresponding spatial optimization strategies for ancient fair public spaces.

2.2. Social Network Analysis Method for Historical Districts

Social network analysis is a critical research method in social and behavioral sciences and involves finite sets of at least one actor and one type of relationships among them [21]. In recent years, it has been applied to historical district conservation.

Regarding network construction, existing studies predominantly treat spatial elements in historical districts as nodes, with transportation connections, functional associations, and behavioral trajectories between nodes serving as connections. Zhou H. Y. et al. [10] identified 19 spatial nodes based on the daily behavioral paths of residents and spatial usage frequencies along Pingjiang Road in Suzhou, constructed a space network and a “resident–space” relational network, and revealed residents’ preferences for spatial nodes. Yu C. et al. [11] treated 71 public spaces in Dashilan as nodes, built a daily life network based on residents’ daily activity trajectories, explored spatial network characteristics and social segregation patterns in Dashilan Historic District, and proposed public space structural optimization strategies.

In terms of network analysis, related studies compare network characteristics by constructing different networks and analyzing indicator results. Chen J. H. et al. [12] developed public space and public life networks for Yuanxia Village; identified mismatches between spatial and activity networks—through comparisons of density, clustering coefficient, node degree, and betweenness centrality—and proposed micro-renewal strategies. Song J. et al. [13] constructed spatial and behavioral networks for Mochou Village, analyzed network features—using density, average path length, and λ-set metrics—and employed QAP regression to identify influencing factors. Xiao Y.T. et al. [14] established public space and rural social networks for Dongxing Village, analyzed social network structures—through degree centrality, clustering coefficient, and modularity metrics—revealed the social relationships and activity demands of villagers, and evidenced support for public space renovations.

Employing the social network analysis method and correlated optimization strategies in historical district, Sepideh Zarepour Sohi et al. [15] employed SNA integrated with the SWOT framework and identified Tehran’s historic bazaar’s core issues in using social media data, prioritizing comfort, vitality, and safety. Their design-driven solutions emphasized walkability enhancement, activity diversification, and natural surveillance reinforcement to optimize spatial quality. Shi Y. L. et al. [16] analyzed SNA metrics (density, k-core, degree centrality) to reveal contrasting network structures in Chongqing’s historic districts—flat areas exhibited stable monocentric networks, while topographically fragmented mountainous zones formed decentralized polycentric configurations. They proposed terrain-responsive strategies involving tiered public space allocation and adaptive spatial interventions to address structural disparities.

Although SNA effectively constructs and analyzes network relationships, existing research demonstrates that discussions on the causes of mismatches remain insufficient, requiring integration with complementary methodologies. Therefore, this study constructs public space and public activity networks to analyze network matching characteristics through relative metrics, combining field investigations and in-depth interviews to explain the causes and reveal existing issues, finally proposing targeted optimization strategies.

3. Case Study

The Gongming Ancient Fair, situated in Gongming Community, Gongming Subdistrict, Guangming District, Shenzhen, covers approximately 50,000 square meters. Bounded by Xiagong Road and Minsheng Avenue to the north, Zhenxing Road to the south, and Renmin Road to the east (Figure 1), it is adjacent to multiple industrial zones and urban villages and represents a typical original-preservation ancient fair. Due to conflicting interests that hindered regional economic development, Gongming Fair was established in 1929 at the junction of Heshuikou and Shangcun Village, modeled after Yuen Long Fair in Hong Kong, with the intention of emphasizing equitable trade. It was officially named Gongming Fair in 1931. Once established, it served as Guangming’s sole commercial center and held market days on the 2nd, 5th, and 8th days of the lunar calendar, attracting traders from Huangjiang, Dongguan, with peak foot traffic reaching 500–600 daily. After the 1980s, over 90% of its arcades were classified as dangerous structures due to aging and fire hazards, with its subsequent decline prompting merchants to relocate to new markets. In 2004, Gongming Ancient Fair was listed as an immovable cultural heritage preservation site in Baoan District. While historically associated with intangible cultural heritage elements such as lion dance and Cantonese opera, these traditions remain unrepresented at the site.

Jiefang Street, the primary thoroughfare of Gongming Ancient Fair, stretches 204.5 m north–south and 91.6 m east–west. The northern section of Jiefang Street retains the only century-old arcade building in Shenzhen, which is government-leased for exhibition purposes and periodically open to the public. The southern section is flanked by two-story shops predominantly engaged in general merchandise, groceries, daily necessities, and textiles. It also preserves the Gongming Granary, a historical structure in Gongming that functions as a material warehouse for electronics factories. Other historical and cultural elements, except for a sealed ancient well relic under conservation, remain unused, and their potential historical significance has not been realized (Figure 2).

4. Methodology

First, residents’ profiles and public activity patterns in Gongming Ancient Fair were collected through questionnaire surveys. Second, public activity paths and spatial nodes were obtained through path tracking and field surveys to construct public activity and space networks. Seven metrics (node degree, density, betweenness centrality, betweenness centralization, clustering coefficient, average path length, and small-world quotient) were analyzed to identify the network characteristics and nodal positions of public spaces and activity networks.

4.1. Data Sources

4.1.1. Questionnaire Survey

The study questionnaire was designed by referring to relevant scholars’ research [22,23] and incorporating Gongming Ancient Fair’s characteristics as well as preliminary survey findings. Based on street-grid demographic data, the ancient fair’s registered population totaled 603 individuals, with non-local residents accounting for approximately 77%. Using Calculator.net for non-scaled questionnaire analysis, a sample size exceeding 58 ensures a reliability level above 0.8. A total of 63 questionnaires were distributed, with 61 valid responses, ensuring sample reliability. Further, 75% of respondents were middle-aged or young adults. Most had relocated to the area due to occupational needs, with residency durations exceeding five years. Regarding public activities, 46% of respondents engage in leisure activities between 6:00 p.m. and 10:00 p.m., highlighting the significant demand for evening exercise. Fitness facilities, children’s playgrounds, and rest spaces ranked highest among the desired spatial improvements, indicating that existing public spaces fail to meet residents’ needs for public activity, family recreation, and relaxation (Figure 3 and Figure 4).

4.1.2. Path Tracking

To minimize essential activities’ influence on the research outcomes, the survey periods were at 8:00, 10:00, 12:00, 15:00, and 18:00 at weekends. Improving on one-on-one tracking methods by simultaneously monitoring multiple activity paths, drones recorded five-minute public activity videos during each time slot. A total of 25 activity paths were extracted from the footage to construct the public activity network.

4.1.3. On-Site Photography

Aerial drone photography and field observations documented crowd activity clusters across different time periods within the research area, providing foundational data for categorizing public space elements.

4.2. Public Space Node Selection

Micro-public spaces, as community-level public spaces, play a vital role in fostering neighborhood cohesion, stimulating social vitality, and promoting social integration from a socio-residential perspective [24]. Micro-public space perspectives define the public spaces in Gongming Ancient Fair as outdoor areas with observable human activity. Totally, 20 of these spaces were categorized into four types based on their locations and typologies: shop-front public spaces, residential-front public spaces, micro-plaza public spaces, and educational and medical facilities (Figure 5 and Figure 6).

4.3. Network Metric Selection

4.3.1. Node Degree

Node degree refers to the number of nodes directly connected to a given node. As node degrees vary significantly across networks of different scales, direct comparisons between networks are challenging. Therefore, the relative node degree is commonly analyzed instead, calculated as the ratio of a node’s degree to the total degrees of all network nodes.

4.3.2. Density

Density reflects the overall connectivity of a network. Higher density indicates stronger inter-node connections and better network cohesion, while lower density implies weaker associations and greater node isolation.

4.3.3. Betweenness Centrality

Betweenness centrality measures the frequency at which a node lies on the shortest path between other node pairs, reflecting its role as an intermediary. A higher value indicates greater mediating influence, signifying a node’s centrality within the network [14].

4.3.4. Betweenness Centralization

This metric describes the disparity between the highest betweenness centrality value and other nodes in the network. A larger disparity implies higher centralization, indicating that the network may comprise multiple subgroups overly reliant on a single node for connectivity, thereby highlighting its critical role.

4.3.5. Clustering Coefficient

The clustering coefficient represents the average of all nodes’ clustering coefficients in the network, reflecting the overall clustering cohesiveness. A higher value signifies tighter node groupings and stronger connectivity.

4.3.6. Average Path Length

This metric calculates the mean of the shortest topological distances between any two nodes in the network, indicating global connectivity. Shorter average path lengths denote better node accessibility and network integration, while longer lengths imply dispersed node distribution and poor reachability.

4.3.7. Small-World Propensity

A small-world network exhibits high connectivity and accessibility, typically measured by the small-world quotient. A larger quotient indicates stronger small-world characteristics. Networks are considered small-world-prone when the quotient exceeds 1:

$$SW=\left(C_{actual}/L_{actual}\right)/\left(C_{random}/L_{random}\right)$$

(1)

where

• $C_{actual}$: Clustering coefficient of the actual network.
• $L_{actual}$: Average path length of the actual network.
• $C_{random}$: Clustering coefficient of a random network of equivalent size.
• $L_{random}$: Average path length of a random network of equivalent size.

5. Results

This study employed SNA to construct a public space network and public activity network of Gongming Ancient Fair, analyzed their compatibility, and explored the causes of network mismatch, thereby providing a basis for recommended optimization strategies targeting the holistic network and its node spaces.

5.1. Public Space Network Construction and Data Analysis

Each node in the public space network represents a public space. Two public spaces are considered connected if they are mutually accessible without traversing any other public space. The matrix construction follows this connectivity principle, assigning “1” to connected node pairs and “0” otherwise. The matrix was subsequently imported into the software Ucinet 6.755 for analysis. The Gongming Ancient Fair public space network metrics reveal the following characteristics:

The network has a density of 0.211, indicating weak inter-nodal connections and limited spatial interactions. The low betweenness centralization suggests relatively balanced nodal statuses and minor subgroup formation. The clustering coefficient of 0.373 reflects dispersed spatial node distribution and weak nodal linkages. The average path length of 2.453 implies that reaching one node from another requires traversing three intermediate nodes, demonstrating moderate spatial accessibility. The small-world quotient (SW) of 1.501 (>1) indicates small-world tendency and suggests theoretically favorable connectivity and accessibility, which appears contradictory to the other metric results (

Table 1. Public space network metrics of Gongming Ancient Fair (source: author Liu H.).

Node Count	Relative Average Node Degree	Density	Average Betweenness Centrality	Betweenness Centralization	Network Clustering Coefficient	Average Path Length	Small-World Quotient (SW)
20	4	0.211	13.800	19.44%	0.373	2.453	1.501

). According to Formula (1), SW > 1 occurs when random networks exhibit lower clustering coefficients and average path lengths. This confirms the inherent small-world characteristics of Gongming Ancient Fair; yet, its actual average path length and clustering coefficient remain suboptimal due to road disorder, preventing full realization of the latent connectivity potential.

Using the betweenness centrality values of nodes from

Table 2. Betweenness centrality of public space network nodes in Gongming Ancient Fair (source: author Liu H.).

Node 1	Node 2	Node 3	Node 4	Node 5	Node 6	Node 7	Node 8	Node 9	Node 10
4.843	34.000	45.379	11.400	1.667	43.136	21.333	2.000	0.000	1.367
Node 11	Node 12	Node 13	Node 14	Node 15	Node 16	Node 17	Node 18	Node 19	Node 20
12.036	20.293	6.243	2.583	3.833	10.719	16.764	0.000	10.102	28.302

as node weights and assigning colors to nodes based on spatial categories, the network data were visualized using the software Netdraw 2.179 (Figure 7). Node 3 (Household Utensils Store) exhibits the highest betweenness centrality, while Node 18 (Gongming Granary) displays the lowest value. This indicates that the Household Utensils Store exerts the strongest control over the public space network, acting as a crucial bridge with a central position in the network structure. In contrast, the historically significant Gongming Granary occupies a peripheral location, reflecting its limited intermediary role in spatial connectivity.

5.2. Public Activity Network Construction and Data Analysis

Based on the public space network, a matrix was constructed using pedestrian flow between nodes. This matrix was binarized using the software Ucinet to derive metric data for the public activity network (

Table 3. Public activity network metrics of Gongming Ancient Fair (source: author Liu H.).

Node Count	Relative Average Node Degree	Density	Average Betweenness Centrality	Betweenness Centralization	Network Clustering Coefficient	Average Path Length	Small-World Quotient (SW)
20	4.900	0.258	11.300	24.03%	0.498	2.189	1.782

).

The Gongming Ancient Fair public activity network metrics reveal the following characteristics:

The public activity network density of 0.258 indicates a relatively limited scope in residents’ activity paths. The clustering coefficient of 0.498 suggests a moderate concentration of public activities. The average path length of 2.189 indicates that reaching one node from another requires traversing three intermediate nodes, demonstrating moderate spatial accessibility. The small-world quotient of 1.782 (>1) indicates that the public activity network possesses small-world potential, necessitating further efforts to unlock latent connectivity and integrity. Using betweenness centrality values from

Table 4. Betweenness centrality of public activity network nodes in Gongming Ancient Fair (source: author Liu H.).

Node 1	Node 2	Node 3	Node 4	Node 5	Node 6	Node 7	Node 8	Node 9	Node 10
2.074	15.532	12.415	12.543	3.476	50.342	9.214	19.817	0.533	0.000
Node 11	Node 12	Node 13	Node 14	Node 15	Node 16	Node 17	Node 18	Node 19	Node 20
21.952	11.549	2.750	1.583	3.583	13.987	17.385	0.000	8.150	19.113

as node weights, a betweenness centrality-weighted graph of the public activity network was generated (Figure 8). Node 6 (Utensil Wholesale Store) exhibits the highest betweenness centrality, while Node 18 (Gongming Granary) shows the lowest, indicating that the Utensil Wholesale Store exerts the strongest control over public activities, whereas the Gongming Granary fails to attract crowd interactions. On cross-referencing the public space network results, the only historic building within the ancient fair, Gongming Granary, underutilizes its historical value, necessitating prioritized optimization. Additionally, the survey interviews identified Nodes 16 and 17 as primary public activity locations for residents. However, their low actual betweenness centrality reveal a limited capacity to attract resident activities during practical use, necessitating targeted optimization.

5.3. Comparative Analysis of Public Space and Public Activity Networks

In summary, the Gongming Ancient Fair public space nodes exhibit fragmented spatial connectivity, while public activities demonstrate dispersed distribution. Nodes with betweenness centrality differences exceeding 10 between the two networks were filtered (

Table 5. Nodes with betweenness centrality differences exceeding 10 between the two networks (source: author Liu H.).

Betweenness Centrality	Node 2	Node 3	Node 7	Node 8
Public Space Network	34.000	45.379	21.333	2.000
Public Activity Network	15.532	12.415	9.214	19.817

) and categorized into two types: nodes with higher betweenness centrality in the public space network (Nodes 2, 3, and 7) and nodes with higher betweenness centrality in the public activity network (Node 8). The former indicates that these nodes occupy relatively critical network positions but lack sufficient crowd attraction, thus requiring focused attention in subsequent renovations; the latter suggests that although positioned at network peripheries, these nodes exhibit high crowd attraction potential, necessitating further investigation into the underlying causes.

5.4. Supplementary Field Interviews and Surveys

Field interviews and surveys were conducted to investigate the causes of the nodes’ performance, with the majority of interviewed residents being middle-aged or young females. The causes are as follows: Node 2 (Xin Guang Pastry Shop) only sells pastries during specific annual festivals, with limited food offerings at other times, resulting in minimal activity. Node 3 (Household Utensils Store) features a wide facade and shallow depth, with merchandise spilling out visually, creating a sense of oppression that discourages residents from non-necessary purchases. Node 7 (Teahouse) is rarely visited, and the interviews reveal that residents prefer staying at home or traveling on weekends rather than visiting the teahouse. The newly built spaces at Nodes 16 and 17, adjacent to the century-old arcades, lack supporting facilities and consequently host minimal resident activities or stays. Further, some mothers expressed a desire for parent–child interaction opportunities in the space but found existing facilities inadequate. Afflicted by management deficiencies from the subdistrict office, Node 18 (Gongming Granary) features disorderly enclosed spaces perceived as monotonous by residents, leading to minimal public activity. One respondent noted that the expansive public areas, potentially suitable for community activities like square dancing or cultural performances, remain underutilized due to factory encroachment. Node 6 (Utensil Wholesale Store), with orderly arranged affordable and refined merchandise in its street-side setback space, attracts residents to linger and browse within the setback area. According to the interviews, residents regard Node 8 (Furniture Store) solely as a transit route with no inherent appeal.

6. Discussion

In summary, the existing issues in Gongming Ancient Fair include poor connectivity and accessibility of the overall public space network, as well as mismatched node spaces between the public space network and the public activity network. The interviews and supplementary research findings indicate that it is necessary to optimize the spatial layout in conjunction with the needs of residents.

6.1. Holistic Optimization of Public Space Network

As crucial connectors between spatial nodes, roads significantly influence the connectivity and accessibility of the public space network. Therefore, optimizing the overall public space network requires road adjustments. Based on current conditions, roads in Gongming Ancient Fair are categorized into demolition-designated roads and retention-designated roads. The former refers to roads requiring removal of illegal structures or construction barriers, and the latter denotes those needing standardized traffic paths’ organization (Figure 9).

The matrix was reconstructed and imported into Ucinet (

Table 6. Public space network metrics of Gongming Ancient Fair after road adjustments (source: author Liu H.).

Node Count	Relative Average Node Degree	Density	Average Betweenness Centrality	Betweenness Centralization	Network Clustering Coefficient	Average Path Length	Small-World Quotient (SW)
20	8.500	0.447	6.300	7.40%	0.650	1.663	1.340

). Compared with the unadjusted public space network data (Figure 10), the adjusted public space network shows an increased network density and clustering coefficient, indicating enhanced overall network connectivity; the average path length is 1.663, meaning any two nodes can be reached within two steps and demonstrating improved network accessibility. Both the average betweenness centrality and betweenness centralization decreased, reducing the nodes occupying controlling positions and presenting greater node status equilibrium across the network. With a small-world quotient of 1.340 (>1), the adjusted network exhibits small-world tendencies and stronger global connectivity. In conclusion, road adjustment and standardization measures effectively improve the accessibility and balance of the public space network.

6.2. Renovation of Key Node Spaces

As discussed, the public space Nodes 3, 6, 16, 17, and 18 require focused attention in Gongming Ancient Fair. Based on their selection criteria, these five nodes were categorized into three types: nodes selected based on betweenness centrality (Nodes 3 and 6), nodes identified through field surveys and interviews (Nodes 16 and 17), and nodes distinguished by architectural features (Node 18).

Following the road adjustments (

Table 7. Comparison of node betweenness centrality before and after road adjustments in Gongming Ancient Fair (ranked by magnitude) (source: author Liu H.).

Before Road Adjustments				After Road Adjustments
Node Name	Betweenness Centrality	Node Name	Betweenness Centrality	Node Name	Betweenness Centrality	Node Name	Betweenness Centrality
Node 3	45.379	Node 19	10.102	Node 6	18.326	Node 8	4.818
Node 6	43.136	Node 13	6.243	Node 20	13.844	Node 12	3.636
Node 2	34.000	Node 1	4.843	Node 17	13.401	Node 1	3.092
Node 20	28.302	Node 15	3.833	Node 10	10.953	Node 16	2.848
Node 7	21.333	Node 14	2.583	Node 11	9.156	Node 4	2.090
Node 12	20.293	Node 8	2.000	Node 2	8.273	Node 5	1.965
Node 17	16.764	Node 5	1.667	Node 19	7.776	Node 15	1.400
Node 11	12.036	Node 10	1.367	Node 7	7.700	Node 13	0.943
Node 4	11.400	Node 9	0.000	Node 18	7.633	Node 14	0.734
Node 16	10.719	Node 18	0.000	Node 3	7.412	Node 9	0.000

), Node 3’s control capability was diminished, while Node 6 maintained a high level of control. Thus, Node 3 was excluded from the prioritized nodes whilst Node 6 remained a key focus (Figure 11).

6.2.1. Renovation of High-Control-Capability Node Spaces

Node 6 (Utensil Wholesale Store), situated in a central position of the Gongming public space and exhibiting the highest crowd activity attraction, serves as a critical node. Its renovation should focus on reorganizing internal exhibition spaces and enhancing setback space design along the street. By leveraging its locational advantage, collaborative efforts with local organizations could be initiated to host diverse activities, thereby further stimulating public engagement.

6.2.2. Renovation of High-Expectation Node Spaces

Node 16 and Node 17 primarily exhibit a mismatch between the childcare needs of middle-aged and young women with children and the fitness demands of sports-oriented residents, resulting in spatial configurations that hinder the activities of both populations. Based on the questionnaire findings, sheltered seating and fitness facilities are proposed for installation at Node 16 to fulfill resting and exercise demands, and a miniaturized children recreational space is designated at Node 17 to accommodate childcare needs. According to the survey findings, most residents predominantly utilize outdoor leisure spaces during evening hours. Therefore, night-time illumination is additionally required for both node spaces to facilitate evening activities.

6.2.3. Renovation of Culturally Distinctive Node Spaces

Management deficiencies have led to the underutilization of Gongming Granary and public space encroachment by electronic factories, effectively obstructing activities for nearly all population groups. Consequently, demolishing and relocating factory auxiliary structures and rationally planning traffic paths for both public activities and factory staff are crucial for renovation. Through spatial design (Figure 12), the public space of Gongming Granary will be opened for large-scale activities such as square dancing and ball sports, transforming discrete node spaces into expansive activity environments [25] promoting public activities.

6.3. Enhancement of Public Activity Organization

Successful regeneration requires more than just spatial transformation; it necessitates a reconfiguration of social dynamics and governance structures [26]. The survey results on daily activities indicate that most residents possess personal hobbies and desire making social connections through shared interests, necessitating coordinated efforts by community organizations and local governments to organize public activities addressing diverse demands. Examples form the survey data include children’s programs, community events, outdoor movies, sports, and art exhibitions.

Concerning festive activities, residents’ preference for traditional folk practices reflects cultural identity needs. This prompts governmental and relevant organizations to fully leverage the intangible cultural heritage value of Gongming Ancient Fair, such as by regularly hosting lion dances and Cantonese opera performances. As Shen Y. et al. [27] propose, the creation of culturally distinctive regional spaces should harness the dynamic nature of Intangible Cultural Heritage (ICH) as a medium. By engaging ICH activities to foster human–space interactions integrating individuality and sociality, material spaces acquire cultural and social significance, thereby reinforcing regional identity and laying the groundwork for a cohesive social environment.

7. Conclusions

This study employed social network analysis to construct public space and public activity networks in Gongming Ancient Fair. A comparative analysis of network metrics identified mismatched node spaces between the two networks, with supplementary investigations and interviews further exploring the causes of these discrepancies. The three proposed optimization strategies are as follows: (1) based on spatial conditions, the road network can be optimized from a holistic structural perspective, with Ucinet simulations validating the improved network integrity and connectivity; (2) targeted renovation of prioritized nodes classified through survey and interview findings; and (3) enhanced organization of community public activities. The revitalization of a historically built environment is a complex and conflict-laden issue [28]. By analyzing the relationships between public space and public activities, this study provides sociologically informed guidance for public space revitalization, offering supplemental perspectives for similar historic districts’ regeneration.

One limitation of the drone-based data collection of public activity paths is that structural obstructions potentially compromised pedestrian trajectory accuracy. Future studies should integrate one-on-one tracking survey methods to ensure the precise construction of public activity networks.