合作网络结构特征对创新绩效影响研究

Structural features of cooperative network impact on innovation performance

在学科交融和知识碰撞的时代下,期刊作者合作网络特征对组织内部的创新绩效具有一定影响,其研究结果可以为科研人员组建高效合作者或团队成员提供科学依据,进而尽量避免因感性认识进行成员选择给团队创新绩效造成的负面影响。鉴于数据科学在当今时代的影响力和价值性,文章选择以"数据"为研究主题,收集了2008-2018年期间刊发的17万篇中文期刊论文,通过数据处理获得了766个相对稳定的作者合作网络,解析每个合作网络的结构特征并利用聚类算法对网络结构特征数据进行聚类,进而获得具有近似网络结构特征的簇。与此同时,对具有不同代表性特征的网络簇使用CART决策树算法挖掘和分析其潜在的决策规则,以此来反映不同网络结构特征对团队创新绩效的影响情况。研究结果表明:(1)科研人员应加强合作共赢,尽量避免封闭式的科研发展模式。(2)高绩效科研合作网络内部至少需要一位科研水平较为突出的带头人作为网络内部资源共享和学术交流的枢纽。(3)网络规模的良性扩大也有利于网络内部整体创新绩效的提高。

In the era of interdisciplinary integration and knowledge collision,the characteristics of journal authors’cooperative network have certain influence on the innovation performance within the organization.The research results can provide scientific basis for the researchers to form suitable partners or team members,and thus avoid the perceptual knowledge.The negative impact of member selection on team innovation performance.In view of the influence and value of data science in today’s era,the article chose to use the"data"as the theme,collected 170,000 Chinese journal articles published during 2008-2018,and formed a relatively stable 766 authors through data processing.The cooperative network analyzes the structural characteristics of each cooperative network and clusters the network structure feature data by using clustering algorithm to obtain clusters with approximate network structure features.At the same time,the CART decision tree algorithm is used to mine and analyze the potential decision rules for network clusters with different representative characteristics to reflect the impact of different network structure features on team innovation performance.The research results show interesting knowledge as follows:Firstly,researchers should strengthen cooperation and win-win,and try to avoid closed research development mode.Secondly,one and more leaders with a high level of scientific research needs to be a hub for internal resource sharing and academic exchange within the network.Thirdly,the benign expansion of the network scale is also conducive to the improvement of the overall innovation performance within the network.