基于时间的工作流模型挖掘算法的研究与实现

Research and implementation of time-based algorithm for mining workflow models

提出了一种基于时间的工作流模型挖掘方法以提高工作流模型挖掘技术的准确性.运用事件日志挖掘活动的执行时间和活动之间的直接等待关系、相交关系和后继等待时间,建立活动间的边矩阵与并行活动矩阵,找出边之间的关系,从而得到带有时间性能的工作流模型.给出的实例说明了这种方法在提高模型准确率与描述能力等方面的优势.

To improve mining accuracy of workflow models, a time-based algorithm was proposed. The execution time of tasks and the following wait time, succession relationships, intersection rela- tionships between tasks were discovered by mining event logs. Setting up the edge matrix between tasks and the task parallel matrix,and finding relationships between edges, a workflow model with time factors was generated. An example was presented to demonstrate the improvements on model accuracy and the description ability of the proposed algorithm.