多任务环节的团队合作知识流动网络研究

Research on Knowledge Flow Network of Team Collaboration in Multi-Task Links

为进一步得出任务背景下团队合作知识流动网络演化的特征,建立了多任务环节的团队合作知识流动网络模型,选用Netlogo软件进行智能体仿真。研究结果表明:增加网络中的合作团队数量,增大网络的集聚系数,减小网络的平均路径长度,提高团队间的知识交流频率能够促进网络的知识流动。网络演化过程中,团队平均连接数量先增后减,每个团队的平均连接数量与任务的合作规模呈正相关,与网络的邻域阈值大小,以及团队自身的知识能力无直接关系;关系强度与网络的平均路径长度呈负相关;知识流动收益与平均路径长度呈指数形式增长;小世界效应能够帮助组织识别网络中的核心团队,拉近团队间的社会距离,促进动态核心能力的形成与提升。

In the knowledge economy, many organizations usually disassemble workflow and divide tasks in order to produce knowledge in the form of team-work. This production mode has obvious positive effect on knowledge flow and knowledge creation in organizations. This paper establishes the knowledge flow network model of multi-tasks by improving BA model and extending relevant theories on social network. We use NetLogo to do agent-based simulation, and analyze the characteristics of knowledge flow＇s network revealed in multi-tasks in order to interpret the various impacts of different indices on network＇s income. Section 1 is model establishment. To adapt to the actual condition of teamwork in tasks, we propose relevant research hypotheses, and use the BA model to （1） establish scale-free network model of knowledge flow, （2） make some improvement to the model according to the actual conditions in multi-tasks teamwork, （3） take deductions, and （4） obtain 4 characteristic variables of network. Four variables, including degree distribution, network cluster coefficient, average path length, and knowledge flow＇s frequency, are used as the conditions of simulation design. Section 2 is simulation design. Considering the agent characteristics of knowledge flow in teamwork, we choose NetLogo to do the simulation. To make the simulation more specific and targeted, this paper divides the simulation design into agent design and simulation context design, in which agent design includes design on agent＇s capabilities as well as behavior. Simulation context design contains basic context of network and scale. In agent＇s capabilities design, to differentiate factors like function and characteristic of different teams we use the triple 〈CAE〉 in kene to define every team. Each dimension represents capabilities, abilities and expertise. In designing agent＇s behavior principles, to maintain network evolution＇s dynamic we find three aspects of the connecting principles of agents, including basic connecting principle, updated connecting principle, and principle of knowledge stock and network income. In designing network＇s basic context, to have more intuitive view of network we assign and standardize values to each team＇s triple 〈CAE〉. In designing network＇s scale, to compare the impacts of different number of teams on network＇s evolution we make the initial network＇s nodes as 50, 80 and 100 separately. We transform the relevant equations into NetLogo＇s language and provide inputs to it. In addition, we calculate the number of connected nodes to measure the connecting density as well as the update connecting level. Section 3 is simulation experiment. According to the pre-test result, when setting the tick in the simulation as 50, we obtain different forms of knowledge flow＇s network based on different teams＇ number, as well as respective variations of the absolute and relative indicators. Results of the research show that in order to improve the knowledge flow in the network the simulation has several countermeasures, such as increasing the cooperative teams＇ number in the network, enlarging the clustering coefficient of the network, reducing the average path length of the network, and improving the knowledge communication＇s frequency among teams. In the evolution process of the network, the average connecting number of teams first increase then reduce. Every team＇s average connecting number has a positive correlation with tasks＇ cooperative scale, whereas there is no direct relationship with network＇s neighborhood threshold, or the knowledge level of the team itself. There is negative correlation between relationship strength and the average path length. There is an exponential increase between knowledge flow＇s income and average path length. The small-world effect could help organizations identify the core team in the network, close the social distances among teams, engage the formation, and improve dynamic core competence. This research addresses relevant managing suggestions from two levels, namely the view of team and the view of organization. This paper provides new thoughts and theoretical reference for the deepening development of knowledge flow.