An approach of process similarity measurement based on activity occurrence relationship

DOI编号  10.7641/CTA.2020.90768
2020,37(9):2011-2019

 作者 单位 E-mail 李东月 安徽理工大学 1244321663@qq.com 方欢 安徽理工大学

针对现有的流程模型行为相似性度量方法大都基于变迁的发生序列, 存在时间复杂度较高的问题, 提出一种基于活动发生关系的流程相似性度量方法. 首先, 形式化左右集概念, 并引入违背度和左右集权重$\alpha,\beta$, 通过活动间的关系变化来确定$\alpha$和$\beta$的取值, 进而得到活动的相似性; 然后, 将流程中活动的相似性进行归一化, 提出基于活动发生关系的流程相似性算法(AOR), 进行业务流程的相似性度量. 最后, 使用实际事件数据案例对本文方法进行了评估, 实验结果表明: 该方法能够正确计算两个不同流程之间的相似性, 且具有较好的时空复杂度.

Aiming at the problem of high time complexity of the existing process model behavior similarity measurement methods, which are mostly based on the occurrence sequence of transitions, a kind of process similarity measurement method based on activities occurrence relationship is proposed. Firstly, the concept of left and right sets are formalized, and the degree of violation and the weights of left and right sets $\alpha,\beta$ are introduced. According to the transformation relationship between activities, the values of $\alpha,\beta$ are determined, and the similarities of activities are obtained. Then, the similarities of all activities in the processes are normalized, and further the processes similarity algorithm (named AOR) based on activities occurrence relationship is presented, which is used to measure the similarity of business processes. Finally, an actual event data case is used to evaluate the proposed method. The experimental results show that the proposed method can correctly calculate the similarity between two different processes, and has better space-time complexity.