过程挖掘算法研究综述  

A Survey of Process Mining Algorithms

作　　者：敬思远[1,2] JING Siyuan(School of Artificial Intelligence,Leshan Normal University,Leshan Sichuan 614000,China;University Key Lab in Sichuan Province of Internet-based Natural Language Intelligent Processing,Leshan Normal University,Leshan Sichuan 614000,China)

机构地区：[1]乐山师范学院人工智能学院,四川乐山614000 [2]互联网自然语言智能处理四川省高校重点实验室(乐山师范学院),四川乐山614000

出　　处：《乐山师范学院学报》2020年第12期39-48,共10页Journal of Leshan Normal University

基　　金：互联网自然语言智能处理四川省高校重点实验室开放项目“基于GPU集群的过程大数据挖掘方法研究”(INLP201903);乐山师范学院培育计划项目“基于通用GPU的工作流挖掘算法研究”(ZZ201822)。

摘　　要：首先对过程模型的三种表示方法Petri网、因果网和过程树进行了介绍。指出了目前过程挖掘算法研究中需要注意的三类关键问题。对主流的过程挖掘算法进行了分类总结。深入比较了当前最有效的三种过程挖掘算法Evolutionary Tree Miner、Inductive Miner和Hybrid ILP Miner,并指出了三种算法的优势和不足。最后,对过程挖掘算法的研究趋势进行了总结。Three representation methods of process model,including petri net,causal net and process tree,are firstly introduced.Three key problems that should be noticed in the research of process mining algorithm are indicated.The major algorithms of process mining are summarized in category.Furthermore,three most effective algorithms of process mining,including Evolutionary Tree Miner,Inductive Miner,Hybrid ILP Miner,are compared,and meanwhile the advantages as well as the disadvantages are pointed out.Finally,the research trends of the process mining algorithms are concluded.

关 键 词：过程挖掘 PETRI网 因果网 过程树 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]