一种基于逻辑Petri网的过程挖掘方法  

A Method of Process Mining Based on Logic Petri Nets

作　　者：杜玉越[1] 朱鸿儒[1] 王路[1] 刘伟[1] 

机构地区：[1]山东科技大学山东省智慧矿山信息技术重点省级实验室,山东青岛266590

出　　处：《电子学报》2016年第11期2742-2751,共10页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61170078;No.61472228);山东省泰山学者建设工程专项经费;山东省自然科学基金(No.ZR2014FM009);青岛市科技计划基础研究项目(No.13-1-4-116-jch);山东省优秀中青年科学家科研奖励基金(No.BS2015DX010)

摘　　要：逻辑Petri网是抑制弧Petri网和高级Petri网的抽象和扩展,可在过程挖掘中简洁准确的表示活动之间复杂的业务逻辑关系.本文在传统Petri网挖掘方法的基础上,为了进一步提高复杂系统挖掘模型的简洁度和拟合度,尤其是对并行活动间存在复杂与或关系的系统,提出了一种基于逻辑Petri网的过程挖掘方法,并给出了逻辑Petri网中逻辑变迁的挖掘算法.它可以充分挖掘活动之间的业务逻辑,并且业务逻辑可用逻辑表达式表示.通过与相应Petri网模型的实例比较分析,例证了本文方法的正确性和有效性,且逻辑Petri网模型更加适合日志行为.Logic Petri nets are the abstraction and extension of the Petri nets with inhibitor arcs and high level Petri nets, and can describe the Business Logic among activities concisely and accurately in process mining. To further improve the simplicity and fitness of the mining models of complex systems, especially the systems with the complex AND-OR relation in parallel activities, a method of process mining is proposed based on Logic Petri nets to improve Petri net models in this paper. An algorithm of mining logic transitions is presented to transform the business logic among activities in event logs into logic expressions adequately. Experiment results illustrate that the mining logic Petri nets can represent the event logs more properly and succinctly than the corresponding Petri nets.

关 键 词：过程挖掘 PETRI网 逻辑Petri网 逻辑变迁 挖掘算法 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]