网络结构、企业特征与产学研合作创新能力演化趋势

基于1998—2019年A股上市公司产学研联合专利申请数据，构建动态专利合作网络，采用复杂网络分析方法揭示上市公司产学研合作整体结构特征，并从企业要素密集度、生命周期和区域3个视角探析产学研合作创新中的合作紧密性、知识吸收能力、知识控制能力与集群发展能力。结果发现：无论企业特征如何，随着时间推移，产学研合作更加广泛，紧密性逐渐降低，小世界现象愈发明显，无标度结构趋于稳定，创新集群越来越独立；3种视角下企业知识吸收能力、知识控制能力差距逐渐缩小，集群发展能力差距扩大；与要素密集度相比，生命周期、区域特征对产学研合作网络结构的影响更显著。结论可为激发企业创新活力、促进区域创新协调发展的相关政策制定提供实证支持。

Network Structure,Enterprise Characteristics and Evolution Trends of Enterprise-University-Institute Collaboration Innovation Ability

Based on data of patents application about enterprise-university-institute collaboration associated with A-share listed companies from 1998 to 2019, dynamic patents collaboration networks are constructed.Using complex network analysis methods, we analyze the evolution trends of the whole structure of E-U-I patent collaboration, and explore dynamic patterns of collaboration intensity,knowledge absorptive capacity, knowledge control capacity, and cluster development capacity from three kinds of perspectives which are factor intensity, life cycle, and regional location. The results show that,regardless of enterprise characteristics,the collaborations among enterprises, universities, and institutes become more widespread, density gradually decreases, the small-world properties get greater, the scale-free structure becomes gradually stable and innovation clusters are more and more independent. Gaps in knowledge capacity or knowledge control capacity between enterprises get narrower, but those in cluster development capacity become wider. Compared with factor intensity, life cycle and regional location have a greater impact on the structure of the E-U-I collaboration network. These findings provide empirical support for government decisions to stimulate enterprise innovation vitality and promote the coordinated development of regional innovation.