基于协作行为的半自动人员分配策略

Semi-automatic staff assignment strategy based on collective behavior

为了对业务流程进行手工分配以找出合适的工作人员,提出一种新的方法,该方法半自动地进行业务过程人员分配,在很大程度上减少了手工分配的工作量。新方法基于谱机器学习框架,同时考虑活动的前序关系和工作者的协作行为,这两种关系隐含着活动的相关属性,能够取得较好的人员分配效果。新方法可以学习各种类型的活动,基于业务过程事件日志为每一个执行者分配活动。经过新方法训练所得的模型,可以实现新活动的半自动人员分配。实验证明所提新方法在两个制造企业数据集上可以获得非常高的精确度。

To find the suitable staff by manual assigning the business process,a semi-automatic staff assignment method which could decrease the workload of monitors based on a novel spectral machine learning framework was proposed.Based on machine learning framework,the precedence relationships of activities and the collective behaviors of actors were both considered by the new method.The new method could learn all kinds of activities and allocate activities to each actor based on process event log.The new model obtained by training could realize the semiautomatic staff assignment of new activities.The experimental results showed that the high prediction accuracy with the proposed method on the real-world datasets of two manufacturing enterprise applications was obtained respectively.