科技创新团队的知识网络构建与知识测度研究

Knowledge Network Construction and Knowledge Measurement of the S&T Innovation Team

科技领域交叉融合创新的趋势下,团队合作成为重要的知识生产形式。准确测度团队知识、刻画团队知识结构,是理解团队创新行为机制的基础,对创新团队的管理与发展具有重要意义。本文创新性地结合网络分析方法和知识嵌入视角,对传统的知识测度方法进行改进,使之更适用于科技团队知识测度。首先,使用最大连通子图从全领域合作网络中识别科技创新团队,基于直接关联与间接引用拓展团队成员的个体知识网络,继而集成得到团队加权知识网络。其次,基于网络分析的两大研究视角——位置取向与关系取向,分别构建知识重合度、知识多样性与知识凝聚性三个指标,用于测度团队知识;在指标测度中对以往方法加以改进,将团队知识嵌入全领域知识网络中计算。最后,选择生物医药领域进行实证,对团队知识测度指标及其内涵进行讨论分析。位置取向的知识重合度与知识多样性分别测度共有知识和异质知识,能够反映团队知识的组成成分;关系取向的知识凝聚性则反映团队知识之间的结构一致性和内容连贯性。多数科技创新团队中知识节点的位置分布均衡,但知识之间关系的结构差异较大。知识多样性与创新绩效呈现倒U形关系,是团队创造力尤其是突破性创新能力的重要来源;令人意外的是,知识凝聚性与团队规模正相关,但对创新绩效影响较小。

As integrated innovation is increasing in the Science and Technology(S&T) field,teamwork has been identified as an important way to recreate knowledge.Measuring and portraying the knowledge base of S&T innovation teams with precision is of essential significance for interpreting team behaviors and innovation mechanisms,and for managing and developing these teams.This study creatively combines network analysis and knowledge embeddedness to improve traditional knowledge measures and enhance their suitability for S&T innovation teams.Firstly,we applied Maximal Connected Subgraph to identify S&T teams.Thereafter,we expanded the individual member knowledge network base on direct connection and indirect citation,and then collectively calculated a fine-grained team knowledge network.Further,from the positional and relational approaches,two analysis perspectives in network study,we constructed Knowledge Overlap,Knowledge Diversity,and Knowledge Cohesion to measure team knowledge.When calculating,we improved the previous measurements by embedding team knowledge into the whole field-wide knowledge network.Finally,we chose biomedicine as an empirical case where we further discussed and analyzed the team knowledge measurement indicators and their connotations.Positional indicators,Knowledge Overlap and Knowledge Diversity,measure shared knowledge and heterogeneous knowledge respectively,which reflects the composition of team knowledge;the relational indicator,Knowledge Cohesion,reflects the structural consistency and content coherence of knowledge from different teams.In most teams,knowledge nodes are evenly distributed,but the structure of knowledge differs considerably.An inverted“U”relationship can be observed between Knowledge Diversity,which is an important source of team breakthrough creativity,and team performance.Surprisingly,Knowledge Cohesion proved to have a weak effect on performance,despite exhibiting a positive relation with team size.