APP下载

基于行为轮廓的模块化流程模型修复方法

2021-07-19盛梦君方贤文邵叱风

盛梦君 方贤文 邵叱风

摘 要:模型修复是业务流程领域中新的优化方法。为了让流程模型对系统的描述更加准确,使模型修复方法系统化,且进一步降低成本,减少以往修复算法的迭代次数和时间复杂度,可运用基于行为轮廓的模块化修复新方法,由分解、过滤、选择、修复、组合五个模块构成。首先,使用极大分解算法对模型进行分解,使流程模型呈片段状,便于筛选出问题片段;然后,使用过滤算法对所有片段进行线性过滤从而缩短通过行为轮廓来选择劣子模型的时间,再使用感应挖掘算法对劣子模型进行修复;接着,再对修复后的劣子模型和优子模型进行组合,从而使得修复后的模型能够重放事件日志,且能够与原始模型尽可能保持行为相似性,最后通过实验分析验证了所提方法的可行性。

关键词:模型修复;过程挖掘;行为轮廓;分解;线性过滤

中图分类号: TP391.9文献标志码:A

文章编号:1672-1098(2021)02-0080-07

收稿日期:2020-11-10

基金项目:国家自然科学基金资助项目(61402011,61572035);安徽省自然科学基金资助项目(1508085MF111,1608085QF149);安徽省高校自然科学基金资助项目(KJ2016A208)

作者简介:盛梦君(1996-),女,安徽合肥人,在读硕士,研究方向:perit网理论。

Modular Process Model Repair Method Based on Behavioral Contours

SHENG Mengjun,FANG Xianwen,SHAO Chifeng

(School of Mathematics and Big data, Anhui University of Science and Technology, HuainanAnhui232001, China)

Abstract:Model repair is a new optimization method in the field of business process.To make the process model describe the system precisely, and to systemize the model repair method , cut the cost and reduce the iteration times and time consumption of the previous repair algorithm, a new modular repair method based on model decomposition and behavioral contours may be applied, which consists of five modules of decomposition, filter, selection, repair and combination. Firstly, the maximum decomposition algorithm is used to decompose the model into fragments, easy to filter out the problem fragments. Then, the filtering algorithm is used to filter all fragments linearly therefore shortening the time of selecting the inferior sub model, which will be repaired by applying the induction mining algorithm, with the behavioral contours. Finally, the repaired inferior sub model and the superior sub model are combined to make the repaired model able to replay the event log (i.e. the consistent log) and maintain the behavior similarity with the original model as much as possible. The feasibility of the proposed method is verified by experimental analysis.

Key words:process model repair; process mining; behavioral contours; decomposition; linear filtering

模型修复是一种新的过程挖掘技术,它将事件日志和流程模型作为输入,通过对日志进行分析来发现流程模型中的偏差,进而对模型进行修复。目前,可以通过事件日志来研究系统的真实情况,运用过程挖掘技术得到流程模型。但是,由于挖掘出能够确切描述现实生活的系统模型是困难的,且流程往往会在系统的生命周期中发生变化。……

登录APP查看全文