Research on the Impact of Inter-Industry Innovation Networks on Collaborative Innovation Performance: A Case Study of Strategic Emerging Industries

Abstract

As global economic competition intensifies, collaborative innovation in strategic emerging industries has become a key factor in promoting economic growth and business development, attracting widespread attention. To delve into the mechanisms of collaborative innovation among China’s strategic emerging industries, this study employs a social network analysis. It investigates the characteristics of these industries by analyzing 8,775,496 invention patents, exploring the impact of inter-industry innovation networks on collaborative innovation. Furthermore, this study incorporates industry knowledge acquisition as a mediating variable and the network density within individual industries as a moderating variable, to more comprehensively explain this impact mechanism. We find that relational and structural embeddings within inter-industry innovation networks significantly positively affect collaborative innovation performance, highlighting the importance of partnerships and the critical role of network configurations. Additionally, the breadth and depth of knowledge acquisition mediate the relationship between relational embedding in inter-industry innovation networks and collaborative innovation performance, emphasizing the pivotal role of knowledge acquisition in fostering collaborative innovation. Lastly, the network density within individual industries positively moderates the relationships between both relational and structural embedding in inter-industry innovation networks and collaborative innovation performance, revealing specific impacts of both internal and external industry innovation network characteristics on collaborative innovation. These findings not only provide practical guidance for collaborative innovation across industries but also offer new insights and implications for policy-making and academic research. In future industrial development, special emphasis should be placed on establishing and maintaining partnerships, optimizing inter-industry innovation networks, and enhancing the breadth and depth of knowledge acquisition to foster sustainable development of collaborative innovation. This is crucial for enhancing corporate competitiveness, creating more job opportunities, and driving innovative economic development.

1. Introduction

As the global economy becomes increasingly complex and competitive, industrial collaborative innovation has emerged as a key factor driving economic growth and enterprise development [1,2]. Industrial collaborative innovation can be broadly classified into two categories: intra-industry collaborative innovation and inter-industry collaborative innovation. Intra-industry collaborative innovation refers to the cooperation and innovation activities among different entities within the same industry, such as companies, research institutions, and universities. Inter-industry collaborative innovation, on the other hand, involves cooperation between different industries, achieving technological, resource, and informational complementarity and synergy through cross-industry collaboration [3,4,5]. This type of cooperation not only brings new opportunities and resources to different industries but also effectively promotes innovation and enhances competitiveness [6]. As one of the key drivers of modern economic development, strategic emerging industries play an important role in shaping economic structure, driving technological progress, and leading future development directions [7]. Therefore, in-depth research on the innovation networks among strategic emerging industries has both theoretical value and practical implications for policy-makers and enterprise managers, aiding in promoting high-quality economic development.

Inter-industry innovation networks refer to the relational networks formed during the innovation process among various enterprises (including suppliers, manufacturers, customers, etc.), research institutions (universities, research institutes), government agencies, and other relevant organizations from different industrial sectors [8,9]. According to social network theory, the structure of inter-industry innovation networks can significantly impact collaborative innovation performance between industries. Firstly, the density and centrality of inter-industry innovation network structures directly affect the efficiency and breadth of knowledge flow. In high-density networks, information can spread rapidly, and the tight connections between nodes facilitate knowledge sharing and the advancement of collaborative innovation. For instance, in a highly central network, enterprises located at the core can access more resources and innovation opportunities, thereby playing a leading role in collaborative innovation. Secondly, structural hole theory emphasizes that nodes acting as bridges between different sub-networks can leverage their unique positions to acquire and integrate information and resources from different networks, thereby promoting cross-industry innovation collaboration [10,11]. Such network structures can maximize the use of resource and information asymmetry, enhancing innovation efficiency and performance. Therefore, studying the innovation network structures among strategic emerging industries is crucial for improving inter-industry collaborative innovation performance.

Additionally, in-depth research on collaborative innovation among strategic emerging industries involves more than just identifying influencing factors; it requires a comprehensive understanding of this complex relational network. Knowledge acquisition plays a critical role in the impact of innovation network structures on inter-industry collaborative innovation performance [12]. Knowledge-based theory posits that knowledge is a key resource for gaining competitive advantage, especially in highly dynamic and complex strategic emerging industries, where acquiring and integrating external knowledge through innovation networks can affect collaborative innovation outcomes [1,13]. Simultaneously, the moderating effect of intra-industry network density on the relationship between innovation network structures and collaborative innovation performance cannot be overlooked. According to the diffusion of innovation theory, high-density intra-industry networks can accelerate the dissemination of information and resources, enhancing the responsiveness and flexibility of the entire network, thereby strengthening the impact of network structures on collaborative innovation performance [14]. Hence, focusing on the roles of knowledge acquisition and network density in innovation networks is crucial for a comprehensive understanding of the collaborative innovation mechanisms among strategic emerging industries.

Based on the above discussion, the key issues this study aims to address are

(1)

How do innovation network structures among strategic emerging industries affect inter-industry collaborative innovation performance?

(2)

What is the specific role of industry knowledge acquisition in the relationship between innovation network structures and inter-industry collaborative innovation performance?

(3)

How does intra-industry network density influence the relationship between innovation network structures and inter-industry collaborative innovation performance?

This study will use China’s strategic emerging industries as an example, based on the “Strategic Emerging Industries Classification and International Patent Classification Reference Table” released by the China National Intellectual Property Administration. We will retrieve and obtain different strategic emerging industry invention patent data from the National Intellectual Property Administration’s patent database and the Himmpat patent database, totaling 8,775,496 invention patents. Subsequently, based on the IPC distribution of the obtained patents, we will use tools such as Python 3.9 to calculate the industry span of different patents, as well as their citation situations. Finally, using tools such as Python, we will draw complex relationship networks between strategic emerging industries and calculate related network indicators. This study aims to provide new insights and recommendations for industry development and policy-making by deeply investigating the relationship between innovation networks among strategic emerging industries and collaborative innovation performance.

The main contributions of this study are as follows: (1) This study focuses on the construction and characteristics of innovation networks between industries, exploring how factors such as relationship density and diversity of relationship types affect the performance of collaborative innovation. It deeply reveals the establishment and operation mechanisms of inter-industry collaborative innovation networks. (2) This study uses industry knowledge acquisition as a mediating variable and divides knowledge acquisition into two dimensions, breadth and depth of knowledge acquisition, deeply exploring how inter-industry collaborative innovation networks affect knowledge acquisition, thus affecting inter-industry collaborative innovation performance. This will help reveal the mechanisms of knowledge transmission and integration in industry collaborative innovation. (3) This study emphasizes the moderating role of network density within an industry in the process of industry collaborative innovation. It studies how high network density can enhance or weaken the relationship between innovation networks between industries and inter-industry collaborative innovation performance. This helps to more comprehensively understand the mechanisms of successful collaborative innovation, thereby more effectively promoting economic growth and industry development. (4) This study, through an empirical analysis, verifies and deeply explores the relationships in the theoretical framework. By deeply analyzing the actual situation of different strategic emerging industries, a more comprehensive understanding of the complex relationships between innovation networks, knowledge acquisition, and collaborative innovation in China’s strategic emerging industries will be achieved.

2. Literature Review and Hypotheses

2.1. Industry Collaborative Innovation

Industry collaborative innovation, as a key engine of modern economic development, has become one of the significant drivers of economic growth and business competitiveness [15]. From Schumpeter’s (1942) [16] innovation theory to Porter’s (1980) Five Forces model [17], scholars have elucidated the critical role of innovation in industry competition. Collaborative innovation, as a cooperative model, can facilitate knowledge sharing, resource integration, and risk-sharing, enhancing innovation efficiency [18,19]. Studies by Costa (2020) [20] indicate that the influence of partnership relationships on successful innovation is significant. Although industry collaborative innovation and corporate collaborative innovation both fall under the category of collaborative innovation, they have distinct differences and connections. Corporate collaborative innovation typically refers to cooperation within a single industry or across industries to jointly develop new technologies or products, focusing on direct cooperation between companies [21]. In contrast, industry collaborative innovation not only involves cooperation between companies but also includes cooperation with governments and research institutions [22], especially emphasizing inter-industry collaborative innovation through integrating cross-industry resources and technologies for broader innovation and development. This cooperation model not only breaks through traditional industry boundaries, achieving close connections along the industry chain [23], but also promotes resource sharing and technological complementarity between industries, significantly enhancing the breadth and depth of innovation [24].

Additionally, existing research has found that the success of industry collaborative innovation is influenced by multiple factors. Firstly, the selection and management of partners are crucial for the outcomes of collaborative innovation [25]. Companies need to choose suitable partners, establish trust, and effectively manage the cooperation process. Secondly, knowledge sharing and technology exchange are the cores of collaborative innovation [26]. Complementary knowledge and resource integration among partners can enhance innovation efficiency and quality. Moreover, factors such as policy support, organizational culture, and intellectual property protection also impact the results of industry collaborative innovation [27]. Thirdly, the role of technological infrastructure and digital platforms is increasingly recognized as pivotal in facilitating industry collaborative innovation. Digital platforms can provide a shared space for communication, coordination, and knowledge exchange, thus bridging geographical and organizational boundaries. Research by Veile (2022) emphasizes that digital platforms can enhance transparency and trust among partners, further driving collaborative innovation efforts [28]. Moreover, the cultural alignment between collaborating entities is another crucial aspect. Studies by Jin (2024) suggest that cultural compatibility can significantly reduce friction and misunderstandings during collaboration, thereby fostering a more conducive environment for innovation [29]. This includes both organizational culture and national cultural aspects when international collaboration is involved. Furthermore, funding and financial incentives play a significant role in supporting industry collaborative innovation. Government grants, subsidies, and tax incentives can lower the financial barriers to collaboration and encourage more extensive participation in collaborative projects. According to research by Edunjobi (2024), financial incentives not only motivate companies to engage in collaborative innovation but also help sustain long-term partnerships by mitigating financial risks [30].

However, although scholars have extensively explored industry collaborative innovation, existing studies primarily focus on establishing and maintaining partner relationships, with few studies addressing the mechanisms by which inter-industry innovation networks impact collaborative innovation. Future research should aim to fill this gap by exploring how inter-industry networks facilitate knowledge transfer and innovation diffusion across different sectors. Additionally, there is a need to examine the long-term impacts of industry collaborative innovation on economic development and societal well-being, thereby providing a more comprehensive understanding of its benefits and challenges.

2.2. Industry Innovation Networks and Collaborative Innovation Performance

Inter-industry innovation networks focus on the cooperative relationships and connection patterns between different industries. The “node–edge” model emphasizes interactions between companies [31], and the “weak ties” theory highlights the importance of information transmission [32,33]. Thus, a social network analysis provides a tool to analyze industry innovation networks, offering significant support for revealing the cooperative relationships between industries and their impact on innovation. However, understanding the mechanisms by which industry innovation networks influence the efficacy of collaborative innovation in strategic emerging industries still requires further research. Following Granovetter’s (1985) classic analytical framework of network embedding theory, this study categorizes industry innovation networks into relational embedding and structural embedding [34,35]. The innovation networks between industries are crucial for innovation performance because they facilitate cross-industry knowledge exchange and resource integration, thereby enhancing overall innovation capability and competitiveness.

Relational embedding in industry innovation networks emphasizes the cooperative relationships and interactions between partners [36]. Such embedded relationships help facilitate information sharing, resource integration, and knowledge transfer, thereby enhancing the efficacy of collaborative innovation [18,19]. By establishing strong collaborative relationships and trust mechanisms, partners can communicate and collaborate more effectively, reducing friction and uncertainty in the cooperation process [37]. Existing studies show that close relationships and trust between partners can promote cross-organizational knowledge flow and enhance innovation capabilities and performance [38]. Additionally, close relational embedding can also foster the formation of an innovation culture, encouraging partners to jointly pursue innovation goals and address common challenges [39]. Relational embedding in industry innovation networks can enable partners to share innovation outcomes more easily, accelerating the development and commercialization of new technologies, thereby positively impacting industry collaborative innovation performance. Therefore, the following hypothesis is proposed:

H1: 

Relational embedding in industry innovation networks positively affects industry collaborative innovation performance.

Structural embedding in industry innovation networks emphasizes the positions and connection modes of different industries within the innovation network [40]. Different structural embeddings can lead to varying information flows and resource allocations, thereby affecting the efficacy of collaborative innovation Existing research has found a positive relationship [41] or an inverted U-shaped relationship [42] between centrality and connectivity in industry innovation networks and innovation performance. Higher structural embedding helps enterprises access diverse resources and knowledge, strengthening communication and coordination between partners, thereby providing a more favorable environment for collaborative innovation. Therefore, the following hypothesis is proposed:

H2: 

Structural embedding in industry innovation networks positively affects industry collaborative innovation performance.

2.3. Knowledge Acquisition as a Mediating Variable

Inter-industry innovation networks significantly affect knowledge acquisition in industries, particularly through relational and structural embeddings. Relational embedding emphasizes the close connections between partners across industries, which facilitates the flow and sharing of knowledge and information [43], thereby increasing the breadth and depth of knowledge acquisition. Through these dense network connections, organizations within different industries can access and absorb new technologies and ideas from various industries and fields [44], greatly enriching their pool of innovative resources. Structural embedding focuses on an industry’s position within the overall industry network, including its relationships with the network core or periphery [45,46]. Industries at the network core can more effectively control the flow of resources and information due to their advantageous positions, thus deeply acquiring key and core technological knowledge. This not only enhances the industry’s innovation capacity but also aids in the overall technological advancement and competitiveness of the industry. Therefore, the relational and structural embeddings in inter-industry innovation networks are key mechanisms for driving the acquisition of knowledge breadth and depth, playing a decisive role in enhancing industry collaborative innovation performance.

Knowledge acquisition represents the process by which an enterprise or industry absorbs and integrates new knowledge and technology from the external environment. It plays a critical role in promoting innovation, improving performance, and driving industry collaborative innovation [47,48], playing a critical role in fostering innovation, enhancing performance, and driving industry collaborative innovation [49]. Existing research categorizes knowledge acquisition into two dimensions: breadth and depth of knowledge acquisition [50,51]. The breadth of knowledge acquisition emphasizes the ability of businesses to acquire knowledge from multiple different fields, sources, or regions [50]. In industry collaborative innovation, the complementarity of knowledge between partners can facilitate innovation activities, thereby enhancing collaborative innovation performance. Previous studies have found a positive relationship between the breadth of knowledge acquisition and innovation performance [52]. Organizations within different industries, through cross-boundary cooperation and diversified technological exchanges, can acquire knowledge and experiences from various fields, thus providing more resources and potential for innovative activities. Therefore, this study suggests that at the industry level, the breadth of knowledge acquisition also plays a mediating role between the relational embedding of industry innovation networks and collaborative innovation performance. Hence, the following hypotheses are proposed:

H3: 

The breadth of knowledge acquisition mediates the relationship between industry innovation networks and collaborative innovation performance.

H3a: 

The breadth of knowledge acquisition mediates the relationship between relational embedding in industry innovation networks and collaborative innovation performance.

H3b: 

The breadth of knowledge acquisition mediates the relationship between structural embedding in industry innovation networks and collaborative innovation performance.

The depth of knowledge acquisition refers to the ability of an enterprise or industry to deeply explore and master high-level, specialized knowledge and technology within a specific field [50,52]. It not only includes a profound understanding and application of existing knowledge but also involves continuous learning and innovation to maintain a leading position in that field. Previous studies have shown that deep knowledge acquisition can enhance a firm’s innovation capacity and technological strength, enabling it to gain a competitive advantage in specific areas [53,54]. By deeply understanding and mastering knowledge in a particular field, firms can better respond to market demands and competitive pressures, thereby improving collaborative innovation performance [55]. In industry collaborative innovation, deep knowledge acquisition may help accelerate the transformation and commercialization of technology, thus enhancing collaborative innovation performance. Therefore, the following hypotheses are proposed:

H4: 

The depth of knowledge acquisition mediates the relationship between industry innovation networks and collaborative innovation performance.

H4a: 

The depth of knowledge acquisition mediates the relationship between relational embedding in industry innovation networks and collaborative innovation performance.

H4b: 

The depth of knowledge acquisition mediates the relationship between structural embedding in industry innovation networks and collaborative innovation performance.

2.4. Network Density within a Single Industry as a Moderating Variable

Network density within a single industry refers to the intensity of connections among participants within a strategic emerging industry, i.e., the degree of interconnectedness among these participants [56]. High network density implies more internal connections, which may influence the flow of information, cooperation opportunities, and knowledge sharing both within and outside the industry [57]. A dense network can facilitate a faster dissemination of innovations and best practices, thereby enabling participants to quickly adapt to market changes and technological advancements [58]. This is crucial for the collaborative innovation process, as innovation often requires various forms of resources and knowledge. High network density may help improve the accessibility of these resources and knowledge, thereby potentially fostering collaborative innovation [59]. This interconnectedness can lead to synergistic effects where the strengths of one industry complement the weaknesses of another, facilitating more holistic and impactful innovations [60]. However, the impact of network density within an industry is not only internal but may also affect the relational and structural embeddings of inter-industry networks. Specifically, high network density might strengthen the connections and cooperation opportunities both within and outside the industry, thereby enhancing the effects of relational and structural embeddings in inter-industry networks [61]. This could lead to more extensive industry cooperation, richer resource sharing, and more collaborative innovation opportunities, ultimately enhancing collaborative innovation performance. Thus, the following hypotheses are proposed (Figure 1):

H5: 

Network density within an industry will strengthen the relationship between inter-industry innovation networks and industry collaborative innovation performance.

H5a: 

Network density within an industry positively moderates the relationship between relational embedding in inter-industry innovation networks and collaborative innovation performance.

H5b: 

Network density within an industry positively moderates the relationship between structural embedding in inter-industry innovation networks and collaborative innovation performance.

3. Research Methodology

3.1. Sample Selection and Data Sources

The data for this study primarily come from the HimmPat patent database. HimmPat integrates global patent data and translations; it supports a semantic search, advanced search, S system command retrieval, and image retrieval. It enables viewing of up to 100 patents per page with full drawings, application number/group merging, and other efficient browsing methods. The database automatically generates smart guided searches for key information such as patent families, citations, serial applications, applicants, inventors, agents, and classification numbers. Currently, the database has accumulated over 180 million patent records, design patents from 74 offices, and legal status data from 119 offices, offering high-precision semantic searches in any language. The data collection and processing steps for this study are as follows:

Step 1: Using an advanced search in the HimmPat patent database with the search formula “(A OR B OR C OR D OR E OR F OR G OR H OR I)/essi AND apd = 20120101–20211231”, this retrieves all invention patent data applied for by strategic emerging industries from 2012 to 2021, obtaining information such as patent titles, applicants, types of applicants, application numbers, application dates, types of patents, IPC classification numbers, and emerging industry classification numbers (

Table 1. Comparison of different symbols for strategic emerging industries.

Symbol (Broad Category)	Classification Name of Strategic Emerging Industries (Broad Category)	Symbol (Middle Class)	Strategic Emerging Industries’ Classification Name (Middle Class)	Symbol (Broad Category)	Classification Name of Strategic Emerging Industries (Broad Category)	Symbol (Middle Class)	Strategic Emerging Industries’ Classification Name (Middle Class)
A	New-generation information technology industry	A1	Next-Generation Information Network Industry	E	New energy vehicle industry	E1	New Energy Vehicle Manufacturing
		A2	Electronic Core Industry			E2	New Energy Vehicle Devices and Accessories Manufacturing
		A3	Emerging Software and New Information Technology Services			E3	New Energy Vehicle Facilities
		A4	Internet and Cloud Computing, Big Data Services			E4	New Energy Vehicle Related Services
		A5	Artificial Intelligence	F	New energy industry	F1	Nuclear Power Industry
B	High-end equipment manufacturing industry	B1	Intelligent Manufacturing Equipment Industry			F2	Wind Power Industry
		B2	Aviation Equipment Industry			F3	Solar Energy
		B3	Satellite and Application Industry			F4	Biomass and Other New Energy Industries
		B4	Rail Transportation Equipment Industry			F5	Smart Grid
		B5	Marine Engineering Equipment Industry	G	Energy saving and environmental protection industry	G1	Energy-Efficient Industry
C	New materials industry	C1	Advanced Iron and Steel Materials			G2	Advanced Environmental Protection Industry
		C2	Advanced Non-ferrous Metal Materials			G3	Resource Recycling Industry
		C3	Advanced Petrochemical and Chemical Materials	H	Digital creative industries	H1	Digital Creative Technology and Equipment Manufacturing
		C4	Advanced Inorganic Non-metallic Materials			H2	Digital Cultural Creative Activities
		C5	High-performance Fibers and Products and Composite Materials			H3	Design Services
		C6	Frontier New Materials			H4	Digital Creativity and Integration Services
		C7	New Materials Related Services	I	Related services	I1	New Technology and Innovative Entrepreneurship Services
D	Bio-industry	D1	Biomedical Industry			I2	Other Related Services
		D2	Biomedical Engineering
		D3	Bio-agriculture and Related Industries
		D4	Biomass Energy Industry
		D5	Other Biological Industries

). A total of 8,775,496 invention patents were retrieved.

Step 2: Using Python 3.9 software, based on the “Strategic Emerging Industries Classification and International Patent Classification Reference Table (2021) (Trial)”, each patent’s specific strategic emerging industry sub-category is calculated. The relationships and quantities between different strategic emerging industry sub-categories each year are analyzed. Additionally, the number of patents cited and the countries or regions of these citations are calculated annually for different strategic emerging industry sub-categories. Moreover, in subsequent calculations and empirical studies, this research focuses on the sub-categories of strategic emerging industries.

Step 3: This stage involves calculating inter-industry collaborative innovation performance. Since this study represents inter-industry collaborative innovation performance by the number of innovative outcomes produced through collaboration between industries each year, the number of invention patents generated through collaborative innovation annually by different industries is retrieved from the HimmPat database using industry names and dates.

Step 4: This stage involves calculating related network indicators. Based on the relationships and data within and outside the industries, a network or models are constructed using packages such as networkx in Python, and network structure indicators for the industries within strategic emerging industries are calculated annually. These include relational and structural embedding indicators in the innovation network and network density indicators within industries. The overall innovation network for the industries from 2012 to 2021 is shown in Figure 2, illustrating the close connections between industries.

3.2. Variable Explanation

(1)

Independent Variable—Inter-Industry Innovation Network

In this study, the inter-industry innovation network is measured from the perspective of innovation network embedding, specifically including relational and structural embeddings of the inter-industry innovation network. Relational embedding primarily examines the trust and reciprocity levels among industries interconnected within the innovation network, typically measured by the strength of relationships. Strong relationships imply more frequent information exchange and cooperation opportunities among network members, significantly impacting an information search and exchange, knowledge sharing and acquisition, and collaborative development [62]. Relational embedding (INRE) focuses on the network characteristics of industries embedded within cluster networks and their relationships with other industries, representing the strength of relationships among interconnected industries within the innovation network [63]. This study follows Phelps (2010) [64] and others, using “the number of collaborations between the target industry and other industries” as a measure of an industry’s relational embedding. Specifically, the natural logarithm of one plus the number of times the target industry collaborates with other industries (TICOI) is used as the measure of industry relational embedding. The formula is below:

INRE = ln(TICOI + 1)

(1)

Structural embedding refers to the impact of an industry’s relative position within the innovation network on the industry itself. This study follows Jiang (2020) [65] and others, using degree centrality from network metrics to measure an industry’s structural embeddedness, where a higher industry degree centrality indicates a more central position within the network. The formula is below:

C(n_i) = d(n_i)/n − 1

(2)

C(n_i) represents the relative pointwise centrality of node n_i and n is the total number of nodes.

(2)

Dependent Variable—Inter-Industry Collaborative Innovation Performance

Invention patents, as a form of intellectual property, are an important reflection of an industry’s innovation capability and technological strength. This study adopts the approach of Liu (2020) [66], quantifying inter-industry collaborative innovation performance as the number of invention patents resulting from industry collaborative innovation outcomes, specifically patents associated with ≥2 industry types. To reduce heteroscedasticity in the model, this study uses the natural logarithm of the number of invention patent applications produced by inter-industry collaborative innovation as the measure of industry collaborative innovation performance, where the collaborative innovation performance in year t is the natural logarithm of the number of invention patents applied for in year t.

(3)

Mediating Variable—Knowledge Acquisition

Knowledge acquisition plays a critical role in the industry innovation process, encompassing knowledge and technology obtained from external sources, providing valuable resources and momentum for industry innovation. This study follows Ruan (2015) [67], Zhang (2019) [68], and others, dividing knowledge acquisition into two dimensions: breadth and depth. In this study, the natural logarithm of the number of countries or regions citing the industry’s patents is used as the indicator for the breadth of knowledge acquisition. The natural logarithm of the number of citations of the industry’s patents is used as the indicator for the depth of knowledge acquisition.

(4)

Moderating Variable—Network Density within a Single Industry

Network density within a single industry is a key metric that measures the intensity of connections among participants within a specific strategic emerging industry [69]. This metric reflects the density of the internal network, i.e., the degree of interconnectedness and cooperation among participants (

Table 2. Measurement of Variables.

Variable Type	Variable Name	Variable Symbol
Dependent Variable	Collaborative Innovation Performance between Industries	ICIP
Independent Variable	Innovation Network Embedding between Industries	INSE
	Inter-industry Innovation Network Relationship Embedding	INRE
Mediating Variable	Breadth of Knowledge Acquisition	KAB
	Depth of Knowledge Acquisition	KAD
Moderator Variable	Network Density within a Single Industry	IND

). The calculation of industry network density is as follows:

D = 2L/(N·(N − 1))

(3)

where D represents the network density, L represents the actual number of relationships, and N is the total number of participants (nodes). The value of network density ranges from 0 to 1, with values closer to 1 indicating a denser network and tighter connections among participants.

3.3. Model Design

In this study, an OLS mixed-effects model is employed, based on panel data, to analyze the relationships among inter-industry innovation network embedding, knowledge acquisition, and inter-industry collaborative innovation performance within strategic emerging industries. Specifically, inter-industry innovation network embedding is used as the independent variable, inter-industry collaborative innovation performance as the dependent variable, knowledge acquisition as the mediating variable, and network density within a single industry as the moderating variable. The following empirical models are constructed to test the hypotheses:

First, the impact of inter-industry innovation network embedding on inter-industry collaborative innovation performance is modeled as follows:

ICIP_i,t = α₀ + α₁IN_i,t + γYear + ε_i,t

(4)

where ICIP represents the inter-industry collaborative innovation performance; IN includes the relational embedding (INRE) and structural embedding (INSE) of the inter-industry innovation network; i and t represent the industry and year, respectively; and ε_i,t is the residual term.

Second, when exploring the mediating effect of knowledge acquisition, a three-step method is used to test the mediating effect of knowledge acquisition. The specific model is as follows:

KA_i,t = α₀ + α₁IN_i,t + γYear + ε_i,t

(5)

ICIP_i,t = α₀ + α₁IN_i,t + α₂KA_i,t + γYear + ε_i,t

(6)

where KA represents the mediating variable knowledge acquisition, including the breadth and depth of knowledge acquisition; ICIP is the inter-industry collaborative innovation performance; IN includes relational and structural network embedding; i and t represent the industry and year, respectively; and ε_i,t is the residual term.

Finally, when exploring the moderating effect of network density within a single industry, the specific model is as follows:

ICIP_i,t = α₀ + α₁IN_i,t + α₂IN*IND + γYear + ε_i,t

(7)

where IND represents the moderating variable industry network density, IN*IND is the interaction term between network embedding and industry network density, and other variables remain as previously defined.

4. Results

4.1. Descriptive Statistics and Correlation Analysis

Table 3. Descriptive Statistics for Main Variables.

Variable	N	Mean	p50	SD	Min	Max
ICIP	400	7.115	7.341	1.546	1.792	9.906
INRE	400	7.675	7.753	1.865	2.944	12.143
INSE	400	0.025	0.026	0.004	0.009	0.03
KAB	400	3.577	3.689	0.574	0	4.159
KAD	400	11	11.277	1.96	0	13.592
IND	400	0.189	0.188	0.047	0.041	0.353

presents the descriptive statistics for the main variables used in this study, with standard deviations for each variable within normal ranges, indicating minimal influence from outliers. The minimum value for inter-industry collaborative innovation performance is 1.792, with a maximum value of 9.906, and a mean value of 7.115. Industries with notably high inter-industry collaborative innovation performance include the Next-Generation Information Network Industry, Emerging Software and New Information Technology Services, Artificial Intelligence, Intelligent Grid Industry, and New Technology and Innovation Entrepreneurship Services (Figure 3). Looking at the indicators for inter-industry innovation networks, in terms of relational embedding, the minimum value is 2.993, the maximum value is 12.143, and the mean value is 7.675; for structural embedding, the minimum value is 0.009, the maximum value is 0.030, and the mean value is 0.004, indicating significant differences in network embedding levels among industries within strategic emerging industries, which is suitable for a subsequent regression analysis. Regarding the indicators for knowledge acquisition, both the breadth and depth of knowledge acquisition have minimum values of 0, with a maximum value for the breadth of knowledge acquisition at 4.159, and for the depth of knowledge acquisition at 13.592. The correlations among variables show that relational embedding in inter-industry innovation networks, structural embedding in industry innovation networks, the breadth of knowledge acquisition, and the depth of knowledge acquisition all have significant positive correlations with inter-industry collaborative innovation performance (

Table 4. Correlation analysis of main variables.

	ICIP	INRE	INSE	KAB	KAD	IND
ICIP	1
INRE	0.786 ***	1
INSE	0.765 ***	0.580 ***	1
KAB	0.742 ***	0.369 ***	0.622 ***	1
KAD	0.892 ***	0.655 ***	0.733 ***	0.913 ***	1
IND	0.329 ***	0.170 ***	0.334 ***	0.283 ***	0.235 ***	1

t statistics in parentheses. *** p < 0.01.

).

4.2. Impact of Inter-Industry Network Indicators on Collaborative Innovation Performance

Table 5. Regression Results for Relational and Structural Embedding in Inter-Industry Collaborative Innovation Networks and Inter-Industry Collaborative Innovation Performance.

	(1)	(2)	(3)	(4)
	ICIP	ICIP	ICIP	ICIP
INRE	0.652 ***	0.635 ***
	(25.38)	(23.37)
INSE			0.315 ***	0.315 ***
			(23.72)	(26.65)
_cons	2.113 ***	2.010 ***	−0.753 **	−1.455 ***
	(10.42)	(8.31)	(−2.24)	(−4.45)
N	400	400	400	400
Year	No	Yes	No	Yes
R2	0.618	0.623	0.586	0.679
adj. R2	0.617	0.613	0.585	0.671

t statistics in parentheses. ** p < 0.05, *** p < 0.01.

reveals the causal relationships between the embedding indicators of innovation networks in strategic emerging industries and the performance of inter-industry collaborative innovation. Column (1) shows the regression results of relational embedding of inter-industry collaborative innovation networks on performance without adding the year as a dummy variable, while column (2) includes the year as a dummy variable; column (3) presents results without the year for the impact of inter-industry innovation network embedding on collaborative innovation performance, and column (4) includes the year as a dummy variable. From the final regression results of different models, it can be seen that relational embedding and structural embedding consistently have a significant positive relationship with inter-industry collaborative innovation at the p < 0.01 level, validating hypotheses H1 and H2. This indicates that there is a positive correlation between the innovation network indicators of various industries within strategic emerging industries and the performance of inter-industry collaborative innovation.

4.3. Mediating Effect of Knowledge Acquisition

Table 6. Test Results for the Mediating Effects of Knowledge Acquisition Breadth and Depth.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
	ICIP	KAB	KAD	ICIP	ICIP	ICIP	KAB	KAD	ICIP	ICIP
INRE	0.635 ***	0.104 ***	0.691 ***	0.488 ***	0.268 ***
	(23.37)	(6.87)	(16.44)	(27.55)	(13.27)
INSE						0.315 ***	0.949 ***	0.382 ***	0.223 ***	0.117 ***
						(26.65)	(16.30)	(22.50)	(16.53)	(9.77)
KAB				1.415 ***					0.971 ***
				(25.26)					(10.74)
KAD					0.531 ***					0.517 ***
					(28.32)					(21.90)
_cons	2.010 ***	2.659 ***	5.500 ***	−1.753 ***	−0.910 ***	−1.455 ***	1.007 ***	0.734	−2.433 ***	−1.834 ***
	(8.31)	(19.72)	(14.70)	(−8.32)	(−5.27)	(−4.45)	(6.25)	(1.56)	(−8.07)	(−8.36)
N	400	400	400	400	400	400	400	400	400	400
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
R2	0.623	0.150	0.439	0.857	0.877	0.679	0.434	0.587	0.753	0.857
adj. R2	0.613	0.128	0.425	0.853	0.874	0.671	0.419	0.576	0.746	0.852

t statistics in parentheses. *** p < 0.01.

presents the results of the mediation effect test of knowledge acquisition breadth and depth between inter-industry innovation networks and inter-industry collaborative innovation. A hierarchical regression method is used to test the significance of the mediation effects of knowledge acquisition breadth and depth. The first step tests the relationship between the independent variable (inter-industry innovation network) and the dependent variable (inter-industry collaborative innovation performance). Columns (1) and (6) in Table 6 test the relationships for relational and structural embedding, respectively, finding significant positive relationships, suggesting that the embedding of innovation networks among strategic emerging industries enhances inter-industry collaborative innovation performance. The second step tests the relationship between the independent variable (industry innovation network) and the mediating variable (knowledge acquisition). Columns (2), (3), (7), and (8) examine the relationships between relational embedding, structural embedding, and both the breadth and depth of knowledge acquisition, finding significant positive relationships. The third step tests the relationship between the independent variables (relational and structural embedding), the mediating variable (knowledge acquisition breadth and depth), and the dependent variable (inter-industry collaborative innovation performance). According to columns (4), (5), (9), and (10) in Table 6, the results remain significant. Hypotheses H3 and H4 of this study are validated. This indicates that the breadth and depth of knowledge acquisition play a mediating role in the relationship between relational embedding, structural embedding of inter-industry innovation networks, and the performance of inter-industry collaborative innovation.

4.4. Moderating Role of Industry Network Density

To test the impact of internal industry network characteristics on the relationship between inter-industry network structure and inter-industry collaborative innovation performance, this study constructs interaction terms between industry network density and relational (X1T) and structural (X2T) embedding of inter-industry innovation networks, exploring their impact on collaborative innovation performance. This study finds that X1T has a significant positive effect on inter-industry collaborative innovation performance (regression coefficient = 0.796, p < 0.01) and X2T also has a significant positive effect (regression coefficient = 0.1472, p < 0.01) (

Table 7. Test Results for the Moderating Effects of Industry Network Density.

	(1)	(2)	(3)	(4)
	ICIP	ICIP	ICIP	ICIP
INRE	0.635 ***		0.460 ***
	(23.37)		(10.43)
INSE		0.3147 ***		0.2792 ***
		(26.65)		(17.57)
x1T			0.796 ***
			(4.95)
x2T				0.1472 ***
				(3.29)
_cons	2.010 ***	−1.455 ***	2.061 ***	−1.353 ***
	(8.31)	(−4.45)	(8.76)	(−4.17)
N	400	400	400	400
R2	0.623	0.679	0.645	0.688
adj. R2	0.613	0.671	0.635	0.679

t statistics in parentheses. *** p < 0.01.

). Therefore, hypotheses H5a and H5b are validated. This indicates that the higher the network density within an industry, the more it enhances the positive impact of relational embedding and structural embedding of inter-industry innovation networks on the performance of inter-industry collaborative innovation.

4.5. Endogeneity Test

In the research process, to prevent bias caused by omitted variables or measurement errors in the inter-industry collaborative innovation network, which could lead to endogeneity problems, we conducted an instrumental variable (2SLS) test. We selected “industry R&D tax relief” as the instrumental variable. On one hand, industry R&D tax relief is related to the independent variables (inter-industry relational and structural embedding) because high R&D tax relief typically encourages increased R&D investment within industries, promoting inter-industry cooperation and interaction. On the other hand, industry R&D tax relief is determined by government policy and is not directly related to the industry’s innovation performance, making it an effective exogenous instrumental variable. According to the two-stage least squares (2SLS) regression in

Table 8. Endogeneity Test.

Phase I
Variant	INRE	t-Value	p-Value	INSE	t-Value	p-Value
R&D Tax Credits	0.6	7.5	<0.01	0.5	6.8	<0.01
Constant Term	1.5	8.7	<0.01	1.3	8.2	<0.01
N	400			400
R2	0.48			0.43
Phase II
Variant	ICIP	t-Value	p-Value
^INRE	0.7	7	<0.01
^INSE	0.6	6.5	<0.01
Constant Term	2.1	9.5	<0.01
N	400
R2	0.55

, the validity of industry R&D tax relief as an instrumental variable was confirmed. The results show that relational embedding and structural embedding still have a significant impact on inter-industry collaborative innovation performance. Therefore, this study does not have endogeneity issues.

4.6. Robustness Test

To further test the robustness of the conclusions, this study employs four methods: regressing with lagged independent variables of inter-industry innovation networks, winsorizing the data, changing the sample size, and replacing the dependent variable. Firstly, using a lagged variable regression analysis ensures causality to a certain extent. By lagging the inter-industry innovation networks by one, two, and three periods, respectively, the impact of lagged strategic emerging industry innovation network embedding on inter-industry collaborative innovation performance is examined. The results are still significant and robust, and comparing the coefficients of network embedding regression across different lagged years shows a decreasing trend in the impact of relational and structural embedding on inter-industry collaborative innovation performance. Additionally, after trimming the sample data at the 5% bilateral percentile, the results remain robust (

Table 9. Robustness Test Results with Lagged and Winsorized Data.

	Hysteresis Test						Bilateral Indentation Processing
	ICIP	ICIP	ICIP	ICIP	ICIP	ICIP	ICIP	ICIP
INRE							0.676 ***
							(23.53)
L.INRE	0.556 ***
	(22.75)
L1.INRE		0.464 ***
		(18.34)
L2.INRE			0.419 ***
			(12.18)
INSE								0.382 ***
								(22.68)
L.INSE				0.374 ***
				(17.65)
L1.INSE					0.361 ***
					(13.03)
L2.INSE						0.352 ***
						(10.21)
_cons	1.681 ***	1.057 ***	1.405 ***	2.157 ***	2.353 ***	2.055 ***	1.925 ***	2.499 ***
	(5.65)	(4.91)	(5.95)	(4.51)	(6.56)	(7.18)	(8.51)	(5.85)
N	360	320	280	360	320	280	360	360
Year	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
R2	0.487	0.476	0.483	0.426	0.437	0.452	0.582	0.564
adj. R2	0.482	0.471	0.475	0.421	0.431	0.447	0.581	0.563

t statistics in parentheses. *** p < 0.01.

). Next, this study conducts a robustness test by changing the sample size. Due to the over-18-month delay in the granting of invention patents, data from 2012 to 2018 were used for the analysis, and the results were found to be robust. Finally, this study employs a robustness test by replacing the dependent variable. By changing the calculation method of the dependent variable and using the total number of inventions, utility models, and design patents applied for through inter-industry collaboration as a measure of innovation performance, the results were still found to be robust (

Table 10. Results of Robustness Tests for Changing Sample Size and Replacing the Dependent Variable.

	Changing the Sample Size		Replacing the Dependent Variable
	ICIP	ICIP	ICIP	ICIP
INRE	0.446 ***		0.676 ***
	(22.96)		(23.53)
INSE		0.273 ***		0.473 ***
		(19.75)		(21.75)
_cons	1.388 ***	1.028 ***	1.366 ***	1.365 ***
	(6.77)	(4.73)	(3.02)	(5.63)
N	280	280	400	400
Year	Yes	Yes	Yes	Yes
R2	0.491	0.481	0.357	0.238
adj. R2	0.486	0.473	0.354	0.234

t statistics in parentheses. *** p < 0.01.

).

5. Discussion

5.1. The Impact of Inter-Industry Innovation Network Structure on Collaborative Innovation Performance

This study finds that the relational embedding of inter-industry innovation networks has a significant positive impact on inter-industry collaborative innovation performance. This is consistent with previous research results, further proving the importance of establishing strong partnerships in the process of collaborative innovation [70]. Relational embedding emphasizes trust and reciprocity between partners, which helps reduce information asymmetry and cooperation risks, providing a more reliable foundation for innovation within the industry. In this context, different industries can share key technologies and knowledge more confidently, promoting technological exchanges and thereby improving the efficiency and effectiveness of collaborative innovation [71]. For example, in the manufacturing industry, a tight supply chain network can significantly improve the efficiency of collaborative innovation, not only in terms of product development speed but also in resource integration and cost control. By sharing information and resources within the supply chain, industries and their internal organizations can respond to market demands more quickly, reduce R&D and manufacturing costs, and thus improve collaborative innovation performance [72]. Moreover, with the advent of the digital era, social media and online collaboration platforms provide new opportunities for more flexible and extensive relationships between enterprises. These platforms not only promote global collaboration but also accelerate the dissemination of innovative ideas, offering a broader stage for collaborative innovation.

This study also finds that the structural embedding of inter-industry innovation networks similarly has a significant positive impact on inter-industry collaborative innovation performance. Structural embedding refers to the centrality and connectivity of an industry within the innovation network, which makes it easier for the industry to access and integrate diverse resources, including technology, talent, and market information. This integration of resources not only improves innovation efficiency but also enhances the industry’s ability to respond to complex market demands. In high-tech industries, companies that occupy central positions in the innovation network are usually able to acquire new technologies and market information earlier, thereby better grasping market opportunities [73]. This centrality enables these companies to lead joint R&D and joint marketing efforts, further improving the efficiency of collaborative innovation. However, achieving and maintaining such centrality requires continuous effort from the industry and its organizations, including establishing and maintaining good partnerships, actively participating in industry associations, alliances, and exhibitions to expand their influence within the innovation network [74]. Through the above discussion, this study not only verifies the positive impact of the inter-industry innovation network structure on collaborative innovation performance but also reveals the important roles played by relational embedding and structural embedding. These findings provide new perspectives and guidance for future research and practice.

5.2. The Mediating Role of Industry Knowledge Acquisition

This study shows that inter-industry innovation network embedding affects collaborative innovation performance by influencing industry knowledge acquisition. Specifically, industry knowledge acquisition plays a crucial mediating role between relational embedding and structural embedding and collaborative innovation performance. Firstly, relational embedding impacts the breadth and depth of industry knowledge acquisition, thereby affecting collaborative innovation performance. Industries with high relational embedding are more likely to engage in knowledge sharing and information exchange, thereby expanding the breadth and depth of knowledge acquisition. In an environment of relational embedding, industries can widely acquire knowledge and technology from different fields and regions. This broad knowledge acquisition not only expands the industry’s innovation horizons but also enhances its competitiveness and innovation capability. For example, companies within an industry can absorb new technologies and management experiences by sharing knowledge with partners from different industries, thereby improving their innovation performance. This broad knowledge acquisition enables companies within the industry to innovate from multiple perspectives in collaborative innovation, providing diverse support for product and service innovation [75]. Additionally, deep knowledge acquisition also plays an important role in relational embedding. Deep knowledge accumulation allows industries to establish competitive advantages in specific fields and lead innovation directions in cooperation [76]. By deeply researching and accumulating knowledge in specific fields, industries not only enhance their own innovation capabilities but also effectively engage in technical exchanges and joint R&D with partners, thereby improving collaborative innovation performance.

Secondly, structural embedding impacts the breadth and depth of industry knowledge acquisition, thereby affecting collaborative innovation performance. In a highly structurally embedded network, industries can leverage their network position to access and integrate diverse resources, including technology, talent, and market information [77]. Structural embedding makes it easier for industries to acquire diverse knowledge and technology from the network. This broad knowledge acquisition allows companies to absorb cutting-edge technologies and innovative ideas from different fields during the innovation process, thereby enhancing their innovation capability and performance. For example, industries occupying central positions in the network can acquire emerging technologies and market information earlier and apply this knowledge to innovation practices, enhancing their performance in collaborative innovation [78]. Deep knowledge acquisition is also important in structural embedding. By deeply researching and accumulating knowledge in specific fields, companies can occupy key positions in the innovation network and become leaders in knowledge and technology. This deep knowledge accumulation enables companies to lead in collaborative innovation, promote joint R&D and technological innovation among partners through technological leadership and knowledge sharing, and thereby improve overall collaborative innovation performance.

These findings indicate that when promoting industry collaborative innovation, industries should not only emphasize the importance of relational embedding and structural embedding but also enhance their innovation capabilities and competitiveness through broad and deep knowledge acquisition. This provides new perspectives and strategies for industries in formulating innovation strategies and cooperation plans, highlighting the critical role of knowledge acquisition in industry collaborative innovation. Future research can further explore the specific mechanisms of different types of knowledge acquisition on innovation performance and how to optimize relational and structural embedding in innovation networks to maximize the effectiveness of collaborative innovation.

5.3. The Moderating Role of Intra-Industry Network Density

This study finds that the density of intra-industry networks significantly moderates the relationship between inter-industry innovation network embedding (both relational and structural embedding) and collaborative innovation performance. Specifically, industries with high network density not only enhance inter-industry collaboration effectiveness but also further facilitate the integration and utilization of innovative resources, thereby improving collaborative innovation performance. Firstly, a high-density industry network environment significantly enhances the positive impact of relational embedding on collaborative innovation performance. In high-density networks, the links between enterprises are closer, and information flows more smoothly. In such an environment, enterprises within the industry can more easily establish and maintain trust relationships, reduce information asymmetry, and improve the efficiency and effectiveness of cooperation. For example, Hua (2022) found that high-density networks promote the rapid dissemination of knowledge and resources, thereby enhancing innovation capability and performance [79]. In high-density networks, the frequency of interactions between enterprises within the industry is higher, and cooperative relationships are more stable. This close cooperative relationship promotes knowledge sharing and technological exchange, further improving the efficiency and effectiveness of collaborative innovation. Yu et al. (2013) also found that social capital is more easily formed in high-density networks, thereby promoting the creation and sharing of knowledge [80]. In such cases, relational embedding can more effectively promote collaborative innovation performance. Additionally, high network density within the industry enhances reciprocity among enterprises, further strengthening the impact of relational embedding on collaborative innovation performance. Enterprises in high-density networks are more likely to form cooperative tacit understandings and trust relationships, thereby improving the success rate and efficiency of collaborative innovation. This is consistent with Uzzi (1996) embeddedness theory, which emphasizes the impact of social relationships on economic behavior and outcomes [81].

Similarly, high network density within an industry also enhances the positive impact of structural embedding on collaborative innovation performance. In high-density networks, the centrality and connectivity of enterprises within the innovation network are more easily strengthened, thereby more effectively integrating and utilizing network resources. Muller (2019) found that high network density improves the speed of information flow and the level of resource sharing within the network, thereby promoting innovation activities [82]. In high-density networks, enterprises in central positions can better leverage their roles in resource integration and innovation leadership. For example, in a high-density network, enterprises can quickly acquire and apply the latest technologies and market information from the network through frequent cooperation and interactions, thereby improving their performance in collaborative innovation. High network density also enhances the connectivity of enterprises within the network, further strengthening the impact of structural embedding on collaborative innovation performance. Enterprises with high connectivity can more broadly acquire and integrate knowledge and resources from different sources, thereby improving their innovation capability and competitiveness. Uzzi (1996) pointed out that high connectivity promotes cooperation and resource sharing among enterprises, thereby enhancing overall innovation performance [81]. This provides new perspectives and strategies for industries in formulating innovation strategies and cooperation plans, highlighting the critical role of network density in industry collaborative innovation.

6. Conclusions

This study delves into the impact of inter-industry innovation networks on collaborative innovation performance through a social network analysis, revealing the crucial roles of industry knowledge acquisition and network density in this process. The results indicate that relational embedding within inter-industry innovation networks significantly enhances collaborative innovation performance by facilitating information sharing, resource integration, and knowledge transfer. Specifically, strong relational embedding helps industries establish closer cooperative relationships and trust mechanisms, reducing information asymmetry and cooperation risks, thus providing a more stable and reliable foundation for innovation. In such a network environment, different industries can confidently share key technologies and knowledge, promoting technical exchanges and collaborative innovation, thereby improving innovation efficiency and effectiveness. Additionally, this study finds that structural embedding also has a significant positive impact on collaborative innovation performance. Industries with high structural embedding, by strengthening their central position and connectivity within the network, can more easily acquire and integrate diverse resources, including technology, talent, and market information, thus enhancing their innovation efficiency and ability to respond to complex market demands. Industries occupying central positions in the network usually gain new technologies and market information earlier, leading joint R&D and marketing efforts, thereby improving collaborative innovation performance. However, achieving and maintaining such centrality requires continuous effort from the industries and enterprises, including active participation in industry associations, alliances, and exhibitions to expand their influence and connectivity within the innovation network.

This study also reveals the mediating role of knowledge acquisition in the relationship between relational embedding, structural embedding, and collaborative innovation performance. Specifically, relational and structural embedding affect the breadth and depth of knowledge acquisition, thereby indirectly impacting collaborative innovation performance. In networks with high relational and structural embedding, industries are more likely to engage in extensive knowledge sharing and information exchange, acquiring diverse knowledge and technologies from different fields and regions, thereby enhancing their competitiveness and innovation capabilities. Broad knowledge acquisition expands the industry’s innovation horizon, providing more innovation resources and potential, while deep knowledge accumulation allows the industry to establish competitive advantages in specific fields, leading innovation directions in collaborations, and promoting joint R&D and technological innovation through technical leadership and knowledge sharing. Moreover, this study finds that single-industry network density significantly moderates the relationship between inter-industry innovation network embedding (both relational and structural embedding) and collaborative innovation performance. In high-network-density environments, the links between enterprises within the industry are closer, and information flows more smoothly, further facilitating the integration and utilization of innovative resources, thereby enhancing collaborative innovation performance. Enterprises in high-density networks more easily establish and maintain trust relationships, reduce information asymmetry, and improve cooperation efficiency and effectiveness. This finding emphasizes the importance of establishing tight connections and efficient cooperation within the industry, helping to form deep cooperative relationships and resource sharing mechanisms, thus enhancing the entire industry’s innovation capability and competitiveness.

In summary, this study empirically validates the positive impact of inter-industry innovation network structures on collaborative innovation performance and reveals the key mediating and moderating roles of knowledge acquisition and network density in this process. These findings provide theoretical support for inter-industry collaborative innovation and offer new insights for policy-making and practical operations. Future industrial development should particularly focus on establishing and maintaining partnerships, optimizing inter-industry innovation network structures, enhancing the breadth and depth of knowledge acquisition, and emphasizing single-industry network density construction to promote sustainable collaborative innovation development. This is crucial for enhancing corporate competitiveness, creating more job opportunities, and driving innovative economic development.

7. Research Implications

This study aims to explore how to enhance the collaborative innovation capabilities of China’s strategic emerging industries through strengthening partnership relationships, optimizing innovation networks, building internal ecosystems, and deepening knowledge acquisition. These research implications not only contribute to the deepening of theoretical research but also provide guidance and reference for practical industrial innovation. The specific research implications are as follows:

(1)

Strengthening Partnership Relationships to Enhance the Innovation Foundation of China’s Strategic Emerging Industries

Research has found that the relational embedding within innovation networks of China’s strategic emerging industries positively impacts the performance of inter-industry collaborative innovation, highlighting the importance of establishing solid partnership relationships. Firstly, different industries should focus on constructing relatively stable cooperation patterns, reducing information asymmetry and cooperation risks, and providing a reliable foundation for innovation. This requires a careful selection of cross-industry partners, emphasizing the establishment of mutual trust and win–win relationships. Additionally, industries can more flexibly identify potential partners and accelerate knowledge sharing and technical exchanges through digital means such as social media and online collaboration platforms. However, industries need to balance and manage these relationships carefully to ensure the sustainability and durability of cooperative relationships. Including insights into the collaborative innovation performance of all developing countries, developing countries should particularly focus on the complementarity of knowledge and technology when building partnership relationships, leveraging global innovation resources to achieve localized innovation. By establishing diversified partnerships with developed countries and other developing countries, they can effectively enhance their technological levels and innovation capabilities, thereby promoting economic development and social progress.

(2)

Optimizing Industry Innovation Networks to Lead the Development Direction of China’s Strategic Emerging Industries

The structural embedding of inter-industry innovation networks positively impacts collaborative innovation performance, emphasizing that the centrality and connectivity of industries within the network are crucial for successful collaborative innovation. Highly interconnected industry innovation networks enable organizations to obtain new technologies and market information earlier, enhancing the efficacy of collaborative innovation. Organizations within industries should elevate their status in the network by actively participating in industry associations, alliances, and other activities. Digital technologies like big data analytics and Artificial Intelligence can be used to more accurately assess opportunities and challenges, optimize partner selection, and enhance the efficacy of collaborative innovation. The openness and dynamism of industry innovation networks offer more possibilities for innovation, requiring organizations to maintain flexibility and adaptability.

(3)

Co-creating Internal Ecosystems to Drive Sustainable Development

Furthermore, the network density within a single industry can be reflected in establishing internal ecosystems and promoting in-depth cooperation among internal enterprises. By building internal ecosystems, various enterprises can form tight connections and interdependencies, promoting the sharing and exchange of knowledge, resources, and technology. Such deep cooperative relationships help accelerate the incubation and dissemination of innovations, enhancing the efficiency and outcomes of collaborative innovation. Moreover, through deep cooperation among internal enterprises, industries can jointly respond to market competition and technological challenges, enhancing the entire industry’s competitiveness and innovation capacity. Therefore, strengthening the network density within individual industries is one of the key strategies for optimizing industry innovation networks, helping to drive the sustainable development and innovative progress of China’s strategic emerging industries.

(4)

Deepening Knowledge Acquisition to Support Collaborative Innovation in China’s Strategic Emerging Industries

The breadth and depth of knowledge acquisition play mediating roles in collaborative innovation, and expanding the breadth and enhancing the depth of knowledge acquisition are key to improving collaborative innovation performance. By actively participating in industry exhibitions, cross-disciplinary training, and other activities, different industries and organizations within them can expand the breadth of knowledge acquisition. Establishing professional R&D teams and deepening cooperation with research institutions can enhance the depth of knowledge acquisition. Intelligent knowledge management systems and joint R&D platforms can more efficiently achieve the breadth and depth of knowledge acquisition. Different industries can use these means to better leverage their innovation potential in collaborative innovation and improve performance levels.