基于融合特征网和模块网的低频行为挖掘方法

针对流程挖掘过程中忽略低频行为的问题,提出一种基于融合特征网和模块网挖掘低频行为的方法.首先,通过处理有效的事件日志确定通讯行为轮廓关系,并根据日志将特征分为不同模块,重构事件内部行为,挖掘相应的模块网与特征网; 然后,融合特征网与模块网得出完整的流程模型,并通过迭代扩展初始模式得出所有低频模式.实例分析证明,本文提出的方法具有一定的可行性.

Low frequency behavior mining method based on feature nets and module nets

Aiming at the problem that the low frequency behavior is ignored in the process mining process, a method of mining low frequency behavior pattern based on fusion feature network and modular network is proposed in this paper. First of all, by processing effective event log to determine the communication behavior profile, dividing the characteristics into different modules according to the log, reconstructing the internal behavior of the event, thus, the corresponding module network and the feature network are excavated. Then, by integrating the feature network and the modular network, the complete process model is obtained, and iterating the initial mode, all the low frequency models is extended. Examples are given to prove the feasibility of the method.