Research on the Performance of Knowledge Co-Creation of Science and Technology Enterprises Based on IUR Network

Abstract

The current research on knowledge co-creation mostly starts from the perspective of process, studying the process of knowledge co-creation, but there is very little research on the performance of knowledge co-creation. As the carrier of enterprise knowledge co-creation, the industry-university-research network (IUR network) provides a platform for enterprise knowledge co-creation. The purpose of this article is to explore the influence of the centrality of the IUR network on the performance of corporate knowledge co-creation, and the mediating role of corporate absorptive capacity. Technology companies are knowledge-intensive companies and have more knowledge co-creation behaviors. Therefore, this article selects the top 100 technology companies in China’s electronic information industry from 2015 to 2019 as the research sample, and establishes the IUR network based on their cooperative patent data. Our empirical results show that: (1) in the IUR network, the higher the network centrality, the enterprise may have better knowledge co-creation performance; (2) the centrality of the industry-university-research network has a significant role in promoting absorptive capacity of enterprises; (3) the absorptive capacity of enterprises has a complete intermediary effect between the centrality of the IUR network and the knowledge co-creation of technology-based enterprises. This research uses the IUR network to study the performance of knowledge co-creation, which further enriches the related research fields of knowledge co-creation.

1. Introduction

In recent years, the rapid growth of Chinese technology-based enterprises such as electronics, information, new materials, and new energy has greatly improved China’s industrial technology level and promoted the development of the entire national economy. Science and technology are the foundations of a country’s prosperity, and innovation is the soul of the nation and the source of power for enterprise development [1]. Especially for technological enterprises, only continuous innovation can quickly respond to market changes and expand their market position [2]. However, due to the limitations of technological inertia and resource constraints, it is difficult for a single individual or organization to independently carry out exploratory innovation activities. Cooperative innovation has become the basic organizational form for enterprises to enhance their exploratory innovation capabilities [3]. In the era of the knowledge economy, the most fundamental of enterprises’ cooperative innovation is knowledge-based innovation activities [4], that is, knowledge co-creation. In the past, simple knowledge sharing and co-creation between members [5,6,7] and departments [8] of an organization could not support the innovation and development of the entire enterprise, especially exploratory innovation. Enterprises must carry out external knowledge co-creation activities in order to seek breakthrough innovation.

However, the current research on knowledge co-creation is relatively scarce. As of October 2021, there are only 55 relevant documents in the advanced search on the China Knowledge Network with the keyword “knowledge co-creation”. A review of relevant domestic and foreign literature found that knowledge co-creation was first proposed by Prahalad and Ramaswamy [9] to describe the knowledge co-creation between enterprises and customers. Later, Chesbrough et al. [10] believed that open innovation included two basic knowledge processes. That is, the knowledge process from the outside to the inside and from the inside to the outside lays the foundation for knowledge co-creation to be divided into embedded knowledge co-creation and alliance knowledge co-creation. Jinjun Nie et al. [11] further subdivided inward and outward knowledge co-creation into four stages: knowledge sharing, knowledge acquisition, knowledge fusion, and knowledge creation through a case study analysis. At present, the research scope of knowledge co-creation mainly focuses on the relationship between enterprises and users, and users and users in virtual communities. Leonard et al. [12] studied that cooperation between enterprises and users in knowledge creation can better understand the organization’s hidden needs for users; Gibbert et al. [13] believe that the best result of the co-creation of enterprise and user knowledge is to improve the organization’s ability to create innovative products and develop new products. With the increase in the complexity of the environment, the research on knowledge co-creation has also risen from two-to-two to more complex co-creation activities between stakeholders with multi-agent participation in the center of the enterprise. Nonaka et al. [14] believe that in product innovation activities involving multiple entities, the participants use their own knowledge to form deep associations between different knowledge structures in the interaction and sharing of knowledge. This association makes a substantial contribution to the dynamic process of knowledge sharing and generation, thereby realizing the co-creation of new knowledge.

A review of previous research shows that the research on knowledge co-creation mainly focuses on enterprises and users, and the current research focuses on enterprise-centered multi-agent stakeholders, but the relevant research is not in-depth enough. At the same time, previous studies have mainly studied knowledge co-creation from the perspective of process, and few scholars have studied knowledge co-creation from the perspective of results, that is, the performance of knowledge co-creation. Knowledge co-creation performance is the result of knowledge co-creation and represents the final output of participants through a series of collaborative activities. It has a decisive effect on whether participants will continue co-creation activities. This is also the significance of this research.

The most important thing for an enterprise looking to co-create external knowledge is to find a suitable co-creation partner. Universities and scientific research institutes are important subjects of knowledge innovation, have abundant knowledge resources [15], and are the best partners for enterprises to co-create external knowledge. Therefore, it is feasible to use the IUR network formed by enterprises, universities, and research institutes to conduct knowledge co-creation research. In addition, the current problems in the research of the existing academic research institutions are out of touch with the actual situation of the enterprise, unable to accurately guide the innovation practice of enterprises [16,17,18,19]. In addition, the low conversion rate of scientific research results of universities, only about 10%, also proves that the promotion of enterprises and universities, academies of sciences for knowledge co-creation carried out, is necessary and practical. In the industry-university-research network, science and technology enterprises issue knowledge requests based on practical problems, and academic research institutions share theoretical knowledge as a knowledge base. This kind of “corporate knowledge” and “theoretical knowledge” collide and combine in the network [20], and jointly create new knowledge meaning that knowledge co-creates performance. In the process of knowledge co-creation of enterprises and academic research institutions, the enterprise first sends out knowledge requests as the knowledge output terminal, and the academic research institutions receive the knowledge requests for theoretical analysis, and output the theoretical results obtained by the enterprise to accurately guide the enterprise to solve practical problems [21,22,23]. At the same time, at the input end of knowledge, enterprises receive the theoretical results of academic research institutions, conduct practical tests, and complete the actual transformation of theoretical results. That is to say, in the process of co-creation, on the one hand, it solves the problem of the transformation of scientific research results of academic research institutions, on the other hand, it also enables enterprises to solve practical problems with theoretical guidance, at the same time creating new knowledge and skills and achieving innovative development. Therefore, using the IUR network as a platform to study the enterprise’s knowledge co-creation behavior and how to improve the enterprise’s knowledge co-creation performance has important theoretical and practical significance.

On the other hand, the internal competence view of enterprises believes that the ability of enterprises to absorb knowledge (the ability of enterprises to acquire, digest, absorb, and apply knowledge) is the key to enterprise innovation [24]. Sternberg and Arndt [25] believe that the internal absorptive capacity of an enterprise is a more important influencing factor than external relations in the innovation process. Lane et al. [26] believe that enterprise supply networks can provide key resources for enterprise innovation, but these resources can only power enterprise innovation through internalization of absorptive capacity. Based on this, we extend the above-mentioned research [26], incorporate corporate absorptive capacity into the research system of industry-university-research network and knowledge co-creation performance, and jointly study strategies to improve corporate knowledge co-creation performance from the perspective of network theory and of corporate internal capabilities.

In this article, we clarify the relationship between the centrality of the IUR network and the performance of knowledge co-creation. At the same time, it explores the role of absorptive capacity between the centrality of the IUR network and the performance of knowledge co-creation. In addition, the contribution of this article is to carry out relevant research on knowledge co-creation from the perspective of results, instead of continuing the previous process of merely treating knowledge co-creation as an interactive process, and enriching the relevant research on knowledge co-creation. The basis of our analysis is the list of the top 100 enterprises in China’s electronic information industry released by the China Electronic Information Industry Association in 2019. In particular, we used Baiteng.com (https://www.baiten.cn/gjs.html, accessed on 11 January 2020) and the Dawei Patent Search System (http://www.innojoy.com/search/home.html, accessed on 11 January 2020) to search the cooperative patents of these 100 companies. In the end, our sample was composed of 198 companies, universities, and research institutes.

Our main research results indicate that companies that are at the center of the IUR network may have a higher knowledge co-creation performance. In addition, our research has proven that the stronger the absorptive capacity of an enterprise, the better it is for the enterprise to carry out knowledge co-creation and improve the performance of knowledge co-creation, and absorptive capacity has a complete intermediary effect between the network centrality and the performance of enterprise knowledge co-creation.

The paper is structured as follows. Section 2 provides an overview of the background literature. Section 3 describes the data, while Section 4 illustrates the methodology. Section 5 presents and discusses the main results and Section 6 offers some concluding remarks and policy implications.

2. Literature Review

2.1. IUR Network Knowledge Co-Creation Model

Japanese scholars Ikujiro Noaka and Hiroki Takeuchi proposed the famous SECI knowledge transformation model in 1995 [27], and divided knowledge transformation into four stages: socialization, externalization, combination, and internalization. Among them, socialization is the process of transforming personal experience acquired through observation and imitation into shared experience, which is the process of transmitting tacit knowledge; externalization is the expression of tacit knowledge through concepts or icons, which is implicit in the process of transforming knowledge into explicit knowledge; combinatorialization is the fusion of new, existing, and scattered explicit knowledge into a knowledge system, and the process of fusing explicit knowledge; and internalization internalizes the explicit knowledge acquired by individuals or organizations into their own experience and completes the final knowledge creation, which is the process of transforming explicit knowledge into tacit knowledge. The SECI model is widely used by academia to conduct knowledge management-related research. Knowledge co-creation is the advanced stage of knowledge management. It is the continuous fusion and collision of the two parties of co-creation, and finally achieves the cyclic evolution process of knowledge sharing, learning, combination, and innovation. This study combines the four stages of knowledge transformation in the SECI model and believes that knowledge co-creation in the industry-university-research network is a circular chain composed of knowledge sharing, knowledge learning, knowledge combination, and knowledge creation, as shown in Figure 1.

2.2. The Centrality of the IUR Network and the Knowledge Co-Creation Performance

Knowledge co-creation performance is the final result of participants’ knowledge exchange in the network, and it is a measure of knowledge co-creation from the perspective of results. This research defines knowledge co-creation performance as the new knowledge created by network members in the process of knowledge sharing and knowledge collision. For enterprises, this new knowledge may be signed cooperative projects, cooperatively developed new products, jointly applied-for patents, and so on.

Network centrality is a variable that describes the rights of individual actors in the network and symbolizes the ability of individuals to acquire and control network resources [28]. This study defines the centrality of the IUR network as the ability of enterprises to establish connections with other partners in the industry-university-research network and use this connection to obtain resources. The level of network centrality represents the closeness of the company’s relationship [29]. The higher the centrality, the more efficient it is to acquire and integrate more diverse information resources [30].

In previous studies, a lot of literature tried to find the factors that influence the co-creation of knowledge. In the research on the co-creation of corporate knowledge, Maruping and Magni [31] and Tur-nipseed [32] emphasized the positive influence of the organizational innovation atmosphere on individual innovation behavior. That is, the organization’s innovation atmosphere will affect the willingness of customers to participate in innovation, and then affect the effect of knowledge co-creation between customers and enterprises. In the study of knowledge co-creation involving multi-agent participation, Kahn, K.B. et al. [33] believe that the types of participants and the total number of participants will significantly affect the effect of corporate knowledge co-creation. Similarly, Pei Zhang and Ying Yang [34] believe that the scope and depth of participation of the actors will affect the effect of corporate knowledge co-creation. Further, scholars have found that both the number of actors and the degree of participation of actors are closely related to the cooperation network composed of multiple actors. As a result, scholars began to study the influence of network characteristics based on patent cooperation networks and IUR cooperation networks, especially the characteristics of network centrality, on enterprise innovation performance. The research of Ferriani et al. [35] pointed out that members at the center of the network have more advantages in obtaining information resources with higher value, and contacting and understanding companies with development prospects more quickly, so as to obtain opportunities for cooperation and innovation, and adjust the company accordingly. The innovation strategy at this stage has a positive impact on corporate innovation; Zhongchao Wu [36] believes that the centrality of the IUR network is one of the factors that drives corporate innovation. Enterprises with high centrality can have more frequent and close contact with external entities, increasing mutual trust, and trust is the driving force of business cooperation, which can promote knowledge exchange and collision, and ultimately generate new knowledge. Especially in the cooperation between enterprises and universities and research institutes, trust can ensure the authenticity of information exchange, better solve the practical problems of enterprises, and achieve the purpose of co-creation. Mazzola and Perrone [37] believe that companies with a higher degree of centrality can establish collaborative R&D relationships with a large number of network members, have more information sources, have more opportunities for knowledge co-creation, and can generate new knowledge by integrating information from different sources, and then promote the improvement of enterprise innovation performance.

Based on the above discussion, we propose the first hypothesis:

Hypothesis 1 (H1).

The centrality of the industry-university-research network can significantly and positively affect the performance of enterprise knowledge co-creation.

2.3. IUR Network Centrality and Absorptive Capacity

Cohen and Levinthal [38] first proposed the concept of absorptive capacity. Later, Zahra and George [39] defined absorptive capacity based on the dynamic capacity theory and believed that absorptive capacity is a series of practices and processes for enterprises to acquire, digest, transform, and apply external knowledge. In addition, for the first time the absorptive capacity was divided into potential absorptive capacity and actual absorptive capacity. This paper draws on the research of Zahra and George [39], and believes that in the IUR network, absorptive capacity is the process of the enterprise’s acquisition, digestion, conversion, and application of external network knowledge.

As an important way for enterprises to absorb heterogeneous knowledge, the IUR network can provide enterprises with new ideas and abundant external resources [40], which is conducive to improving the absorptive capacity of enterprises. The empirical research of Yong Dai et al. [41] found that the cluster network structure (including network scale, centrality, and strength) has a significant role in promoting absorptive capacity (including potential absorptive capacity and actual absorptive capacity). Empirical research by Jixing Zheng and Jing Liu [42] found that social network structure (including network centrality and network scale) is significantly positively correlated with absorptive capacity. The centrality of the IUR network is of great significance to absorptive capacity. We believe that this is due to the following reasons: first, the more the enterprise is at the center of the IUR network, the more external entities will be connected [43], the more channels there are to acquire rich knowledge resources, enhancing the company’s ability to absorb new ideas. Secondly, the enterprise has increased the knowledge breadth and depth of the enterprise by cooperating with members of the industry-university-research network to share knowledge and resources [44], which is also conducive to improving the absorptive capacity of the enterprise. Once again, enterprises with high network centrality can establish direct and close cooperative relations with network members, which is conducive to understanding the interaction between other members in the network [45]. It can also use close cooperation to better integrate this knowledge and improve the company’s ability to absorb reality.

Based on the above discussion, we thus propose the following hypotheses:

Hypothesis 2 (H2).

The centrality of the industry-university-research network positively affects the absorptive capacity of enterprises.

2.4. Absorptive Capacity and Knowledge Co-Create Performance

According to the view of internal capabilities of an enterprise, the absorptive capacity of an enterprise is a series of processes in which an enterprise acquires, digests, transforms, and applies the external knowledge that it contacts [46,47,48]. Experience has shown that companies with high absorptive capacity have stronger innovation capabilities and it is easier to create new knowledge. Chesbrough [49] believes companies must first fully absorb and integrate external subject knowledge, and then export it to the outside. Therefore, the absorptive capacity of the company can have a positive impact on the innovation performance of the company. Yeoh [50] divides corporate absorptive capacity into potential absorptive capacity and actual absorptive capacity, and believes that potential absorptive capacity is a cross-organizational capacity, and both capacities have a positive impact on innovation performance. This article explains the relationship between absorptive capacity and corporate knowledge co-creation performance from the four dimensions of absorptive capacity, namely knowledge acquisition, knowledge digestion, knowledge conversion, and knowledge application. (1) Knowledge acquisition can enhance the depth and breadth of knowledge of an enterprise [51], and provide a knowledge base for the enterprise to carry out knowledge co-creation; (2) Knowledge digestion: the enterprise can make up for the knowledge gap by digesting and assimilating external knowledge [52] and improving the success rate of co-creation; (3) Knowledge conversion: the process of transforming external homogenous or heterogeneous knowledge and integrating with internal knowledge may generate new ideas and viewpoints [53], which is conducive to the creation of new knowledge by the enterprise; (4) Knowledge application: through the application of knowledge acquired, digested, and transformed in the past, enterprises can truly internalize external knowledge and contribute to the co-creation of knowledge. To sum up, different dimensions of absorptive capacity can positively affect the performance of enterprise knowledge co-creation.

Based on this, this article puts forward the following hypotheses:

Hypothesis 3 (H3).

Absorptive capacity can significantly and positively affect the performance of enterprise knowledge co-creation.

2.5. The Mediating Role of Absorptive Capacity

From the above analysis, it can be seen that the IUR network can provide enterprises with a large amount of valuable information and resources [54], but only through absorptive capacity can enterprises internalize this information into their own resources [26], and then act on the co-creation of corporate knowledge. First of all, although the centrally located enterprises have abundant resources, only the enterprises with a high absorptive capacity can identify and apply the information that is valuable for the enterprise co-creation from the large amount of complicated information [55]. Secondly, in the IUR network, members cooperate based on common interests. Only when both parties can effectively absorb the knowledge shared by each other can the results of knowledge co-creation be maximized and the co-creation performance be improved. That is to say, in the IUR network, only through high absorptive capacity can the company’s positional advantages be effectively brought into play, thereby improving the performance of knowledge co-creation. Previous studies have also proved this point. Escribano [56] and other studies believe that companies can benefit from external knowledge sources, but companies with high absorptive capacity benefit more significantly. Jixing Zheng and Jing Liu [57] found through empirical research on small and micro enterprises that absorptive capacity plays an intermediary role in the social network of small and micro enterprises and enterprise innovation performance.

To sum up, since absorptive capacity is not only the antecedent variable of enterprise knowledge co-creation performance, but also the result variable of the centrality of the IUR network, we believe that enterprises in the center of the IUR network can improve their co-creation performance by improving absorptive capacity.

Based on this, this article puts forward the following hypotheses:

Hypothesis 4 (H4).

Absorptive capacity plays an intermediary role in the centrality of the industry-university-research network and the performance of enterprise knowledge co-creation.

The conceptual model of this study is shown in Figure 2.

3. Data and Descriptive Analysis

3.1. Data Sources

In order to investigate the impact of the IUR network centrality and absorptive capacity on the performance of technology-based enterprises’ knowledge co-creation, we select the top 100 companies in the electronic information industry in China as the research sample. On the one hand, the electronic information industry refers to the industry that uses electronic technology and information technology to produce electronic and information-related products. It is a typical technology-based enterprise with the characteristics of being technology-intensive and knowledge-intensive [58]; on the other hand, the environment in the high-tech field is certainty and rapid product updates put forward new requirements on the knowledge base of the electronic information industry. Therefore, academic research institutions with rich theoretical knowledge have become key cooperation targets for enterprises in the electronic information industry, and it is easy to build an IUR network platform. At the same time, for the electronic information industry, the top 100 companies, as the best among them, are the main force of industry-university-research cooperation. Moreover, the main business income of the top 100 companies accounts for 40% of the entire industry, which is very representative of the entire industry. Therefore, this article chooses the electronic information industry as the research object to build an IUR network, and collects data through the following steps: first, extract the list of the top 100 companies in the electronic information industry in 2019 from the China Electronic Information Industry Association (See

Table A1. Top 100 enterprises in China’s electronic information industry in 2019.

Huawei Technologies Co., Ltd. (Guangdong, China)	Zhejiang Fuchunjiang Communication Group Co., Ltd. (Zhejiang, China)
Lenovo Group (Beijing, China)	Shenzhen DJI Innovation Technology Co., Ltd. (Guangdong, China)
Haier Group (Shandong, China)	Shenzhen Sannuo Investment Holdings Co., Ltd. (Guangdong, China)
Xiaomi Group (Beijing, China)	Wanma United Holdings Group Co., Ltd. (Zhejiang, China)
Peking University Founder Group Co., Ltd. (Beijing, China)	Guangzhou Shiyuan Electronic Technology Co., Ltd. (Guangdong, China)
BYD Company Limited (Shenzhen, China)	Guangzhou Radio Group Co., Ltd. (Guangdong, China)
Sichuang Changhong Electronics Holding Group Co., Ltd. (Sichuan, China)	Shaanxi Electronic Information Group Co., Ltd. (Shanxi, China)
Hisense Group Co., Ltd. (Shandong, China)	Xu Ji Group Co., Ltd. (Henan, China)
BOE Technology Group Co., Ltd. (Beijing, China)	Shanghai Huahong Co., Ltd. (Shanghai, China)
TCL Group Co., Ltd. (Guangdong, China)	Wingtech Communications Co., Ltd. (Zhejiang, China)
Inspur Group Co., Ltd. (Shandong, China)	Pulian Technology Co., Ltd. (Guangdong, China)
Tianneng Battery Group Co., Ltd. (Zhejiang, China)	Zhejiang Jinko Energy Co., Ltd. (Zhejiang, China)
ZTE Corporation (Guangdong, China)	Shanghai Longcheer Technology Co., Ltd. (Shanghai, China)
Hangzhou Hikvision Digital Technology Co., Ltd. (Zhejiang, China)	Jiangxi Helitai Technology Co., Ltd. (Jiangxi, China)
Hengtong Group Co., Ltd. (Jiangsu, China)	Guangdong Shengyi Technology Co., Ltd. (Guangdong, China)
Ziguang Group Co., Ltd. (Beijing, China)	Shenzhen MTC Co., Ltd. (Guangdong, China)
Zhongtian Technology Group Co., Ltd. (Jiangsu, China)	YOFC Optical Fiber and Cable Co., Ltd. (Hubei, China)
China Information and Communication Technology Group Co.,Ltd. (Hubei, China)	Huayu Vision Technology Co., Ltd. (Shanghai, China)
Dongxu Group Co., Ltd. (Hebei, China)	Huike Co., Ltd. (Guangdong, China)
Ningbo Joyson Electronics Co., Ltd. (Zhejiang, China)	Tongling Jingda Special Magnet Wire Co., Ltd. (Anhui, China)
Konka Group Co., Ltd. (Guangdong, China)	Huaxun Ark Technology Co., Ltd. (Guangdong, China)
Tongding Group Co., Ltd. (Jiangsu, China)	Sungrow Power Supply Co., Ltd. (Anhui, China)
OFILM Group Co., Ltd. (Guangdong, China)	Wisdom Haipai Technology Co., Ltd. (Jiangxi, China)
Henan Senyuan Group Co., Ltd. (Henan, China)	Shenzhen Transsion Manufacturing Co., Ltd. (Guangdong, China)
Shanghai INESA Co., Ltd. (Shanghai, China)	Kunshan Liantao Electronics Co., Ltd. (Jiangsu, China)
SMIC International Integrated Circuit Manufacturing Co., Ltd. (Shanghai, China)	Fengfan Co., Ltd. (Hebei, China)
Skyworth Group Co., Ltd. (Guangdong, China)	Oriental Risheng New Energy Co., Ltd. (Zhejiang, China)
Fortis Group Co., Ltd. (Zhejiang, China)	Wuhu Changxin Technology Co., Ltd. (Anhui, China)
Jinglong Industrial Group Co., Ltd. (Hebei, China)	Anhui Tiankang Co., Ltd. (Anhui, China)
Nari Group Co., Ltd. (Jiangsu, China)	Tongguang Group Co., Ltd. (Jiangsu, China)
Huaqin Communication Technology Co., Ltd. (Shanghai, China)	China Silian Instrument Group Co., Ltd. (Chongqing, China)
Tianma Microelectronics Co., Ltd. (Guangdong, China)	Shenzhen Gongjin Electronics Co., Ltd. (Guangdong, China)
Aerospace Information Co., Ltd. (Beijing, China)	Shenzhen Taihengnuo Technology Co., Ltd. (Guangdong, China)
Sunny Group Co., Ltd. (Zhejiang, China)	China Metallurgical CCID Group Co., Ltd. (Chongqing, China)
Shenzhen Stock Exchange Changying Precision Technology Co., Ltd. (Guangdong, China)	United Automotive Electronics Co., Ltd. (Shanghai, China)
Zhejiang Dahua Technology Co., Ltd. (Zhejiang, China)	Camel Group Co., Ltd. (Hubei, China)
Yongding Group Co., Ltd. (Jiangsu, China)	Tianshui Huatian Electronics Group Co., Ltd. (Gansu, China)
Zhenxiong Copper Group Co., Ltd. (Jiangsu, China)	Luxshare Electronic Technology Co., Ltd. (Jiangsu, China)
Shanghai Nokia Bell Co., Ltd. (Shanghai, China)	China Power Tai Chi Co., Ltd. (Beijing, China)
Guangdong Desai Group Co., Ltd. (Guangdong, China)	Leyard Optoelectronics Co., Ltd. (Beijing, China)
Sichuan Jiuzhou Electric Group Co., Ltd. (Sichuan, China)	China Hualu Group Co., Ltd. (Liaoning, China)
Shenzhen Huaqiang Group Co., Ltd. (Guangdong, China)	Hengdian Group East Magnetics Co., Ltd. (Zhejiang, China)
Xinwangda Electronics Co., Ltd. (Guangdong, China)	China Resources Microelectronics Co., Ltd. (Shanghai, China)
Nantong Huada Microelectronics Group Co., Ltd. (Zhejiang, China)	Zhejiang Narada Power Co., Ltd. (Zhejiang, China)
Jiangsu Xinchao Technology Group Co., Ltd. (Jiangsu, China)	Shennan Circuit Co., Ltd. (Guangdong, China)
Tongfang Co., Ltd. (Beijing, China)	China Lucky Group Co., Ltd. (Hebei, China)
GoerTek Co., Ltd. (Shandong, China)	Shenzhen Stock Exchange Xintianxia Group Co., Ltd. (Guangdong, China)
Fujian Electronic Information Co., Ltd. (Fujian, China)	Hytera Communications Co., Ltd. (Guangdong, China)
New H3C Technology Co., Ltd. (Zhejiang, China)	Fuzhou Fuda Automation Technology Co., Ltd. (Fujian, China)
Tianjin Zhonghuan Electronic Information Group Co., Ltd. (Tianjin, China)	Xiamen Hongfa Acoustic Co., Ltd. (Fujian, China)

at the end of the article for the specific list). Then, on the basis of these 100 companies, through the format of “company”, “university”, “company”, and “institute”, the universities and scientific research that have jointly applied for patent patents with Baiteng.com and the Dawei Patent Search System are used to search for them, such as “Huawei Technologies Co., Ltd.” and “Universities”, “Huawei Technologies Co., Ltd.”, and “Institutes”. Finally, collect all the patent data jointly applied by these organizations in 2015–2019 to identify cooperation relations, and establish a network of industry-university-research cooperation. Data selection basis: firstly, enterprises that do not have cooperative patents are excluded; secondly, since the financial data of enterprises are involved in the empirical study later, enterprises that lack financial information are eliminated. Finally, the cooperative network of this study is composed of 198 organizations.

The patent data in this article come from Baiten.com (https://www.baiten.cn/gjs.html, accessed on 11 January 2020) and Dawei Patent Search System (http://www.innojoy.com/search/home.html, accessed on 11 January 2020), and through manual organizing and building the network. Other corporate data are obtained from the corporate annual report of www.cninfo.com.cn/new/index, accessed on 11 January 2020).

3.2. Variable

3.2.1. Result Variable: Knowledge Co-Creates Performance

The performance of knowledge co-creation mainly measures the new knowledge that an enterprise co-creates in the process of collaborative innovation with its partners. For enterprises, this new knowledge may be jointly-developed new products, signed cooperative projects, jointly-applied patents, etc. At the same time, previous studies have proven that patents are an important part of enterprise innovation results, can be used to measure enterprise innovation performance [59], and are more representative than enterprise R&D investment. Therefore, this article selects patent output as a measure of co-creation performance on the basis of previous research, and uses the number of patents jointly applied for by a company and all other partners to represent the company’s knowledge co-creation performance.

3.2.2. Independent Variable: Network Centrality

Commonly used metrics for network centrality include degree centrality, betweenness centrality, and proximity centrality [60], among which degree centrality measures the direct relationship, and betweenness centrality and proximity centrality reflect indirect centrality. Compared with other indicators, the degree centrality represents the number of directly connected members in the network, which can directly reflect the central position of the enterprise [61]. Therefore, this study selects degree centrality as an index to measure the centrality of the network.

3.2.3. Intermediary Variables: Absorptive Capacity

Absorptive capacity is the ability of an enterprise to absorb and digest external knowledge, which is of great significance to enterprise innovation. At present, most studies use indicators such as questionnaire surveys, technological gaps [62], and R&D intensity [63] to measure corporate absorptive capacity. Drawing on the research of Fredrich [63], this paper uses R&D intensity to measure absorptive capacity and calculates it with R&D investment/operating income.

3.2.4. Control Variables

Combined with previous research, this paper selects the following control variables: (1) age of the enterprise (Age) calculated as the year of data collection minus the year of establishment of the company. (2) The size of the company is calculated as the company’s registered capital. (3) Debt to asset ratio (DAR) is calculated as liabilities/total assets.3.3. Variable Descriptive Statistics

Table 1. Descriptive statistics of variables.

	Average Value	Standard Deviation	1	2	3	4	5	6
Knowledge co-creates	327.286	570.396	1
Network centrality	0.046	0.084	0.532 **	1
Absorptive capacity	0.155	0.15	0.641 ***	0.663 ***	1
Age	2.909	0.632	0.02	0.234	0.215	1
Size	18.000	14.659	0.627 ***	0.284	0.441 **	−0.206	1
Debt to asset ratio	0.574	0.108	−0.215	−0.135	−0.141	0.291	−0.26	1

Notes: The table is about the mean, variance, and correlation of variables. ***, ** denote significance of the parameters at the 1%, 5% levels, respectively. Source: Manually sorted it from Baiteng.com and Dawei Patent Search System.

is the descriptive statistics about the variables, including the mean value, variance of the variables, and the correlation between the variables. It can be seen from Table 1 that network centrality, absorptive capacity, and corporate knowledge co-creation performance have a good correlation, which lays the foundation for the subsequent regression analysis. In addition, the VIF values in this study are all less than 3, which is lower than the upper limit of 10, indicating that there is no multicollinearity.

4. Methods

4.1. Construction of IUR Network

In order to explore the relationship between the centrality of the IUR network and the knowledge co-creation performance of technology-based enterprises, this paper first uses the cooperative patent data to construct the IUR network matrix. For cooperative patents, the following selection principles are: (1) the patent applicant must be an enterprise, university, or scientific research institution, not an individual; and (2) the patent applicant must be greater than or equal to 2. That is, the applicant is a combination of companies, universities, and research institutes, not just one. Then, based on the selected cooperative patents, a 0–1 matrix is constructed according to the cooperative relationship between the patent applicants: 1 represents the patents with cooperative applications between the patent applicants, and 0 represents the patents without cooperative applications between the patent applicants. Through the analysis of patent applicants who cooperated in the patent application, a 198 × 198 0–1 matrix is obtained, which reflects the cooperative relationship between the electronic information industry sample matrix and other organizations, and imported the matrix into UCINET6.212 software The network centrality index can be obtained, and the specific calculation path is Netdraw-File-0pen-Ucinet dataset-Network-Centrality-degree. The specific network construction process is shown in Figure 3.

4.2. Empirical Model

After calculating the centrality of the IUR network, this paper constructs a multiple regression model and uses SPSS to perform a regression analysis to verify the hypothesis proposed in Section 2. The reason for choosing multiple regression is that the multiple regression analysis is an analysis method that describes the linear relationship between multiple independent variables and a dependent variable. It can construct a regression equation based on the regression coefficients to predict the value of the dependent variable based on the values of multiple independent variables, and it is more accurate than a single independent variable prediction.

In order to test Hypothesis 1: The centrality of the IUR network positively affects the performance of corporate knowledge co-creation, we constructed model (1):

$${\mathit{Patent}}_{\mathit{it}}{=\alpha }_0{+\alpha }_1{\times \mathit{Degree}}_{\mathit{it}}+{\alpha }_2{\times \mathit{Control}}_{\mathit{it}}+{\epsilon }_{\mathit{it}}$$

(1)

In order to test Hypothesis 2: The centrality of the IUR network positively affects absorptive capacity, we constructed model (2):

$${\mathit{ABC}}_{\mathit{it}}={\beta }_0+{\beta }_1\times {\mathit{Degree}}_{\mathit{it}}+{\beta }_2\times {\mathit{Control}}_{\mathit{it}}+{\epsilon }_{\mathit{it}}$$

(2)

In order to test Hypothesis 3: Absorptive capacity positively affects the performance of corporate knowledge co-creation, we constructed model (3):

$${\mathit{Patent}}_{\mathit{it}}={\gamma }_0+{\gamma }_1\times {\mathit{ABC}}_{\mathit{it}}+{\gamma }_2\times {\mathit{Control}}_{\mathit{it}}+{\epsilon }_{\mathit{it}}$$

(3)

In order to test Hypothesis 4: Absorptive capacity plays a mediating role between the centrality of the IUR network and the performance of knowledge co-creation, we constructed model (4) and compared the results with model (3).

$${\mathit{Patent}}_{\mathit{it}}={\delta }_0+{\delta }_1\times \mathit{Degree}+{\delta }_2\times {\mathit{ABC}}_{\mathit{it}}+{\delta }_3\times {\mathit{Control}}_{\mathit{it}}+{\epsilon }_{\mathit{it}}$$

(4)

Among them, “i” represents the sample company, “t” represents the year, “Patent” represents the knowledge co-creation performance, “Degree” represents the network centrality, “ABC” represents the absorptive capacity of the enterprise, and “Control” represents the control variables, namely the age of the company, the size of the company, and the debt to asset ratio.

5. Results

In Section 4, we constructed four models to test our hypotheses in Section 2. In this section, we use SPSS to perform a regression analysis to verify relevant hypotheses. The regression results are shown in

Table 2. Empirical test results.

	Model 1	Model 2	Model 3	Model 4
	Knowledge Co-Creation	Absorptive Capacity	Knowledge Co-Creation	Knowledge Co-Creation
Age	0.056	0.167	−0.009	−0.023
	(0.773)	(0.399)	(0.958)	(0.898)
Size	0.532	0.316	0.374	0.372
	(0.014 **)	(0.096 *)	(0.055 *)	(0.063 *)
Debt to asset ratio	−0.046	−0.036	−0.038	−0.029
	(0.807)	(0.85)	(0.891)	(0.856)
Network centrality	0.363	0.53		0.11
	(0.074 *)	(0.013 **)		(0.608)
Absorptive capacity			0.546	0.499
			(0.008 ***)	(0.042 **)
R2	0.532	0.531	0.633	0.64
Adjusted R2	0.415	0.414	0.541	0.52
F	4.553	4.534	6.9	5.325
P	0.012 **	0.012 **	0.002 ***	0.005 ***

Notes: This table reports the influence of the centrality of the industry-university-research network and absorptive capacity on the innovation performance of enterprises, as well as the mediating role of absorptive capacity. The centrality of the industry-university-research network is calculated by UCINET 6.212. R² represents an evaluation of the fitting result, which indicates how much of the fluctuation of the Y value can be measured by the fluctuation of the X value. Adjusted R² also considers the sample size and the number of independent variables in the regression, and the representative fitting result is more accurate. In the regression, we use the adjusted R² to represent the goodness of fit. ***, **, and * denote significance of the parameters at the 1, 5, and 10% levels, respectively. Source: Manually sorted it from Baiteng.com and Dawei Patent Search System.

.

5.1. IUR Network Centrality and Knowledge Co-Creating Performance

Model 1 is the regression result of network centrality and corporate knowledge co-creation performance. The regression results show that there is a significant positive correlation between network centrality and corporate knowledge co-creation performance, which is significant at a probability of 10% (regression coefficient 0.363), which is consistent with the research of Mazzola and Perrone [37]. The more centrally located enterprises have more opportunities to contact external resources and higher knowledge co-creation performance. This result shows that the embedded external network of enterprises has a significant role in promoting the co-creation of enterprise knowledge. For technology-based enterprises, occupying the central position of the network makes more use of enterprise innovation and development.

5.2. Centrality and Absorptive Capacity of IUR Network

Like Yong Dai et al. [41], we also study the relationship between the centrality of the industry-university-research network and absorptive capacity. Model 2 is our research result. The regression results are also consistent with Dai Yong et al. [41]. The network centrality and corporate absorptive capacity are significantly positive at a probability of 5%, that is, high network centrality has a positive impact on corporate absorptive capacity. Hypothesis 2 is verified. This result shows that in the IUR network, technology-based enterprises occupy a central position, which is conducive to improving the ability of enterprises to absorb and digest external knowledge.

5.3. Absorptive Capacity and Corporate Knowledge Co-Create Performance

In Model 3, the regression results show that absorptive capacity can significantly and positively affect the performance of enterprise knowledge co-creation, which is significant at a probability of 1% (regression coefficient 0.546). This shows that high-absorptive technology enterprises can significantly improve their knowledge co-creation performance, which supports Hypothesis 3. This is consistent with the research of Chesbrough [49] and Yeoh [50]. That is to say, to improve the performance of knowledge co-creation, enterprises need to improve their own absorptive capacity in order to efficiently absorb internal and external knowledge and achieve the purpose of innovation.

5.4. The Mediating Role of Absorptive Capacity

Model 2 and Model 3 show that changes in the location of the network center significantly affect the same direction changes in corporate absorptive capacity, while changes in absorptive capacity have the same direction changes that can significantly affect the performance of technology-based enterprises’ knowledge co-creation. At the same time, putting the network centrality and absorptive capacity into Model 4, it can be seen that absorptive capacity can significantly affect the knowledge co-creation performance of technology-based enterprises, significant at the 5% level (regression coefficient 0.499), and the influence of network centrality on the performance of knowledge co-creation of technology-based enterprises is no longer significant, which indicates that the absorptive capacity of enterprises completely mediates the influence of network centrality on the performance of knowledge co-creation of technology-based enterprises. That is, the network centrality affects the company’s absorptive capacity and then affects the performance of technology-based enterprises’ knowledge co-creation, which verifies Hypothesis 4 of this article. This is also similar to the research of Escribano [56]. For enterprises, although they can obtain resources from external networks, they can only effectively transform resources if they have a high absorptive capacity. Therefore, an enterprise at the center of the network can only use its own absorptive capacity to digest and absorb external knowledge, in order to give full play to the information advantage of the central position and promote the co-creation of enterprise knowledge.

5.5. Robustness Analysis

Problems such as sample self-selection, sample size, missing variables, etc., may cause endogeneity problems [64], which in turn lead to inaccurate regression results. In order to ensure the reliability of the results, a robustness test is required. This paper selects the performance of knowledge co-creation lagging one period as an instrumental variable and performs the regression analysis again to ensure the reliability of the results.

Table 3. Model robustness test.

	Model 1	Model 2	Model 3	Model 4
	Knowledge Co-Creation	Absorptive Capacity	Knowledge Co-Creation	Knowledge Co-Creation
Age	0.051	0.167	−0.011	−0.027
	(0.794)	(0.399)	(0.951)	(0.885)
Size	0.512	0.316	0.367	0.365
	(0.016 **)	(0.114)	(0.063 *)	(0.071 *)
Debt to asset ratio	−0.055	−0.036	−0.048	−0.038
	(0.771)	(0.85)	(0.774)	(0.825)
Network centrality	0.369	0.53		0.122
	(0.071 *)	(0.013 **)		(0.574)
Absorptive capacity			0.542	0.467
			(0.01 **)	(0.06 **)
R2	0.528	0.531	0.622	0.63
Adjusted R2	0.41	0.414	0.528	0.507
F	4.48	4.534	6.588	5.115
P	0.013 **	0.012 **	0.002 ***	0.006 ***

Notes: This table reports the regression results after replacing the dependent variable with the knowledge co-creation performance that lags behind one period, which is a test of the robustness of the model. R² represents an evaluation of the fitting result, which indicates how much of the fluctuation of the Y value can be measured by the fluctuation of the X value. Adjusted R² also considers the sample size and the number of independent variables in the regression, and the representative fitting result is more accurate. In the regression, we use the adjusted R² to represent the goodness of fit. ***, **, and * denote significance of the parameters at the 1%, 5%, and 10% levels, respectively. Source: Manually sorted from Baiteng.com and Dawei Patent Search System.

shows the regression results of the robustness analysis. It can be seen from the table that after replacing the dependent variable with a one-period lagging knowledge co-creation performance, there is no significant change in the regression results. The influence of network centrality on the performance of knowledge co-creation is significantly positive at the level of 10%, which is consistent with the regression results of the previous article. Therefore, it is sound to prove that the network centrality of the previous article has a significant role in promoting the performance of enterprise knowledge co-creation. At the same time, the influence of network centrality on absorptive capacity is significantly positive at the 5% level, and the influence of absorptive capacity on the performance of corporate knowledge co-creation is significantly positive at the 5% level. By putting network centrality and absorptive capacity into the model at the same time, the effect of the absorptive capacity on the performance of knowledge co-creation is significant at the level of 5%, but the network centrality is not significant. This is consistent with the previous results, indicating that absorptive capacity is related to the performance of network centrality and knowledge co-creation. The result of the intermediary effect is robust.

6. Discussion

This paper mainly studies the performance of knowledge co-creation in the IUR network, using the IUR network as a platform to explore the relationship between network centrality, corporate absorptive capacity, and knowledge co-creation performance. This article is based on the top 100 companies in the electronic information industry, using their 2015–2019 cooperation patents with academic research institutions to build an IUR network. Then, the UCINET6.212 is used to calculate the centrality of the IUR network, and the SPSS is used for the subsequent regression analysis. The research results show that the centrality of the IUR network has a significant positive impact on the performance of enterprise knowledge co-creation. When an enterprise is gradually located at the center of the network, the number of external academic and research entities connected to the enterprise gradually increases, and knowledge co-creation activities become more frequent, which is conducive to improving the performance of enterprise knowledge co-creation. In addition, the absorptive capacity of an enterprise determines the degree of absorption and digestion of resources from a central location, which in turn affects the knowledge co-creation performance of the enterprise. That is, the absorptive capacity of the enterprise plays an intermediary role in the network centrality and the performance of knowledge co-creation. The following are the theoretical contributions and implications of this research for practice, as well as the limitations of this research.

6.1. Theoretical Contribution

Knowledge co-creation is an important subject in the field of open innovation. Previous studies focused on the process of knowledge co-creation [12,13], while ignoring the results of knowledge co-creation. Based on previous studies, this paper uses quantitative methods to extend knowledge co-creation from a process perspective to a result perspective, and analyzes the factors that affect the performance of knowledge co-creation. First, it is proposes that the IUR network formed by school–enterprise cooperation provides a natural platform for knowledge co-creation research, further reveals the influence of the centrality of the IUR network on the performance of enterprise knowledge co-creation, and enriches the quantitative research on the performance of knowledge co-creation. As the complexity of the innovation environment increases, companies rely solely on their own internal knowledge to carry out innovation activities alone, and can no longer adapt to the economic environment of sharing, co-creation, and win–win [65]. Companies need to seek external cooperation at all stages of R&D [66], to expand the scope of corporate knowledge for joint innovation. In the industry-university-research network, information and resources are distributed unevenly and non-linearly in the main body of the network [67]. Different positions represent the difference in information and resources obtained by enterprises in the network. Enterprises in the center of the network occupy more abundant resources. The quantity of resources and higher quality of resources are conducive to the development of knowledge co-creation activities and the improvement of corporate knowledge co-creation.

In particular, we further explore the internal mechanism of network centrality and knowledge co-creation performance, confirming the role of network centrality-corporate absorptive capacity-knowledge co-creation performance. Enterprises at the center of the network have abundant resources, but only when these resources are transformed through absorptive capacity can they truly play a role in the enterprise’s knowledge co-creation. That is to say, although the enterprise can use the external heterogeneous knowledge obtained in the center of the network to act on the enterprise knowledge to co-create performance, this effect is realized through the path of enterprise absorptive capacity.

6.2. Implications for Practice

Co-creation of knowledge is the deepening and development of the field of open innovation, which is of major significance to the innovation and development of enterprises and the country. From a practical perspective, research on the theory and practice of knowledge co-creation can solve the current innovation dilemmas faced by enterprises and provide new ideas for enterprise innovation and development. Our research results also show that companies should fully consider their own development strategies and consider their own information advantages and limitations from a strategic point of view. On the one hand, companies should actively embed the cooperation network, establish a wide range of industry-university-research networks, and evaluate the potential risks of their partners. We will select alliance partners in a targeted manner to form a close cooperative relationship and improve the results of co-creation. On the other hand, for enterprises, occupying a good network position is not the ultimate goal [68]. Enterprises must increase R&D investment, actively exchange knowledge and information with external parties, and at the same time strive to cultivate absorptive capacity, which will transform network resources into the enterprise itself, while improving corporate innovation performance. For the government, it is necessary to formulate relevant policies to encourage enterprises to increase their absorptive capacity, and at the same time increase knowledge exchanges among enterprises, universities, and scientific research institutes to form a good scientific research atmosphere, and ultimately improve the overall absorptive capacity of the region.

6.3. Limitations and Future Research

Our analysis has some limitations, mainly related to the data of empirical analysis, which must be admitted. First of all, this research takes our country’s technology-based enterprises in the top 100 electronic information companies as the research object, but different industries have different cooperation models, which have certain limitations for the conclusions to be extended to other industries, especially traditional industries. In the future, we can try to add data from other industries to conduct further research on the impact of network location and corporate knowledge co-creation performance. Secondly, this study selects the number of cooperative patents to measure the performance of enterprise knowledge co-creation, but cooperative patents cannot fully represent the co-creation results. In the future, data on enterprise cooperation projects and cooperative development of products can be added to construct an index system to measure knowledge co-creation indicators more accurately. Finally, the data collected in this study are time-series data. Time data can see developments and changes, but sometimes they may change the trend due to drastic changes in the environment. In the future, cross-sectional data can be added, and panel data can be used for related research.

7. Conclusions

Starting from the theory of social network and absorptive capacity, this article discusses the influence of the centrality of the IUR network and absorptive capacity on the performance of enterprise knowledge co-creation. In addition, this study also proves that absorptive capacity has a mediating role between the centrality of the IUR network and enterprise innovation performance. The theoretical contributions of this paper are as follows: (1) researching knowledge co-creation from the perspective of results, expanding the research level of knowledge co-creation; (2) taking 198 electronic information industry companies, universities, and scientific research institutions as samples, using a regression analysis, and verifying that the centrality of the IUR network positively affects the innovation performance of enterprises; (3) clarifying the influence mechanism of the IUR network centrality on the performance of knowledge co-creation, which is of great significance to clarify how the IUR network centrality affects knowledge co-creation.