Under the background of implementing the innovation driven strategy and building an innovative country, integrating into the innovation cooperation network is an important way to enhance the regional innovation capacity of China's provinces. Based on the number of papers published in the core collection of web of science between China's provinces from 2008 to 2018, this paper uses the social network analysis method to construct China's provincial innovation cooperation network, and describes its temporal and spatial evolution characteristics from three aspects: network structure characteristics, network density characteristics and agglomerative subgroup analysis, From the perspective of proximity, this paper analyzes the impact mechanism of China's inter provincial innovation cooperation, and tests its robustness. This paper finds that the network structure of inter provincial innovation cooperation in China is becoming more complex and balanced. As time goes on, the links between the network nodes in China's innovation cooperation network are constantly enriched, and more and more node provinces occupy an important position in the network; The structure of China's innovation cooperation network has been continuously optimized, and the regional characteristics of the point degree centrality are obvious. The point degree centrality of each province has been improved in numerical value, and the order of the point degree centrality of most provinces is relatively stable. The middle centrality of each province is decreasing, and the ranking of middle centrality among provinces is changing greatly. Agglomerative subgroup analysis shows that geographical distance is an important factor affecting inter provincial innovation cooperation, and the close level of economic development between provinces is more conducive to innovation cooperation between the two provinces, and the number of agglomerative subgroups shows a downward trend. On the whole, the network characteristics of inter provincial innovation cooperation in China become more obvious, and the level of inter provincial innovation cooperation is constantly improving. This paper tests the impact mechanism of China's inter provincial innovation cooperation from the perspective of multi-dimensional proximity. The econometric test shows that improving network, industrial and economic proximity and shortening geographical distance are conducive to promoting China's inter provincial innovation cooperation, but network proximity and industrial proximity have the greatest impact on China's inter provincial innovation cooperation, while geographical proximity and economic proximity have less impact. Network proximity could influence the cooperation by adjusting geographical proximity and economic proximity. No matter random sampling or substitution econometric model, the robustness test results are consistent. Finally, this paper has some policy implications for the analysis of the characteristics and influence mechanism of China's inter provincial innovation cooperation. Three suggestions are put forward. First, China should focus on creating innovation growth poles in the central and western regions, and promote the balanced development of innovation network pattern. Secondly, each province should put forward ways to optimize innovation cooperation according to its own situation and local conditions. Third, the backward provinces can improve the innovation ecological environment to enhance the status of innovation network. Under the background of implementing the innovation-driven strategy and building an innovative country, integrating into the innovation cooperation network is an important way to enhance the regional innovation capacity of China′s provinces. Based on the number of papers published in the core collection of web of science between China′s provinces from 2008 to 2018, this paper uses the social network analysis method to construct China′s provincial innovation cooperation network, and describes its temporal and spatial evolution characteristics from three aspects: network structure, network density and agglomerative subgroup analysis. From the perspective of proximity, this paper analyzes the impact mechanism of China′s inter-provincial innovation cooperation, and tests its robustness.
    This paper finds that the network structure of inter-provincial innovation cooperation in China is becoming more complex and balanced. As time goes on, the links between the network nodes in China′s innovation cooperation network are constantly enriched, and more and more node provinces occupy an important position in the network. The structure of China′s innovation cooperation network has been continuously optimized, and the regional characteristics of the point degree centrality are obvious. The point degree centrality of each province has been improved in numerical value, and the order of the point degree centrality of most provinces is relatively stable. The middle centrality of each province is decreasing, and the ranking of middle centrality among provinces is changing greatly. Agglomerative subgroup analysis shows that geographical distance is an important factor affecting inter-provincial innovation cooperation, and the close level of economic development between provinces is more conducive to innovation cooperation between the two provinces, and the number of agglomerative subgroups shows a downward trend. On the whole, the network characteristics of inter-provincial innovation cooperation in China become more obvious, and the level of inter-provincial innovation cooperation is constantly improving.
    This paper tests the impact mechanism of China′s inter-provincial innovation cooperation from the perspective of multi-dimensional proximity. The econometric test shows that improving network, industrial and economic proximity and shortening geographical distance are conducive to promoting China′s inter-provincial innovation cooperation, but network proximity and industrial proximity have the greatest impact on China′s inter-provincial innovation cooperation, while geographical proximity and economic proximity have less impact. Network proximity could influence the cooperation by adjusting geographical proximity and economic proximity. No matter random sampling or substitution econometric model, the robustness test results are consistent.
    Finally, this paper has some policy implications for the analysis of the characteristics and influence mechanism of China′s inter-provincial innovation cooperation. Three suggestions are put forward. First, China should focus on creating innovation growth poles in the central and western regions, and promote the balanced development of innovation network pattern. Secondly, each province should put forward ways to optimize innovation cooperation according to its own situation and local conditions. Third, the backward provinces can improve the innovation ecological environment to enhance the status of innovation network.