中文摘要：

为避免制造业企业员工反生产行为（CWB）对组织安全及绩效的负面作用,基于贝叶斯网络（BN）和计算试验方法,研究群体CWB的演化模型。首先将个体CWB发生机制和制造业企业调研数据相结合,通过参数学习构建出员工CWB的BN结构模型;然后基于该模型,结合社会影响和观点交换理论,设计员工群体行为交互规则,并嵌入计算试验模型中进行仿真分析。研究表明：员工规模增长不会导致群体CWB的扩散,而员工尽责性和组织分配公平对CWB的影响显著;员工间沟通范围和地缘关系会导致CWB的恶化,但控制关键在于防范初期的极化舆情;平行管理和有机式组织虽然能控制群体CWB,但在实施时存在风险。

英文摘要：

In order to avoid negative effects of the manufacturing enterprise employees counterproductive work behavior on organizational safety and performance,this paper studies a new way of combing BN with computational experimental method to build on evolution model for CWB.First,based on the individual behavior mechanism of CWB and real data on Chinese manufacturing enterprises,a BN structural model for CWB can be acquired by parameter learning.And then interactive rules are designed for employees group behaviors on the basis of the BN model,social network theory and social exchange theory,which are embedded in the computational experiment model to simulate the CWB evolution process.The research shows that the growth of employees will not lead to the diffusion of CWB,the employees conscientiousness and distributing fairly have a significant influence on CWB,that both the scope of communication and the geographical relationship between employees will lead to deterioration of CWB,but the key to control is preventing initial polarization public opinion,and that while parallel management and organic organization can control group CWB,there are risks in implementation.