Font Size: a A A

Research On Process Mining Methods Based On Regional Event Logs

Posted on:2019-05-30Degree:MasterType:Thesis
Country:ChinaCandidate:W W ShiFull Text:PDF
GTID:2428330545989026Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Process mining is a process model that use the mining method from the event log given by information system.It is also a new technology that can restore the business process.Under the informationized environment,business process management system model for the development of the enterprise is becoming more and more important.Due to the rapid development of information and data,business processes record a large amount of event information.Process mining technology can reproduce the business process,which are good for companies to further study business processes.Therefore,the research of process mining technology is of great significance to the development of business processes.Previous process mining algorithms mainly aimed at fewer event log tasks,mainly based on behavior profiles to determine the dependencies of the event log tasks.Although a few mining process models are based on the process behavior of indirect dependence,it is seldom analyzed from the behavior profile.This paper mainly determines the indirect dependence of event log task through the behavior contour relation of event log task.When the event log have too many tasks or log trails,the previous process mining methods has low mining efficiency and high cost.Therefore,how to mine the process model from a large number of events is very important.Next,how to effectively optimize the process model is also an important problem.In view of these two aspects,this paper build the model based on the research of the process mining method of regional event logs and the optimization of Petri nets.Through the demonstration,it can handle a large number of event log process mining pronlems.The main contributions of this paper are as follows:(1)In view of the problems existing in the previous process mining methods,there are few process models that can mine indirect dependency relationships in event logs.This paper proposes the use of behavioral contour relationships to represent the indirect dependencies of event logs,mainly in Petri nets and Based on the basic knowledge of behavior profiles,a method of process mining based on regional event logs is proposed.Firstly,event logs are divided into different regions based on a large number of event log indirect inheritance matrix values.The regional event log uses an indirect inheritance matrix mining process model.Based on the inactive inheritance process model mining method of active clusters,this paper proposes a process mining method based on regional event logs.Based on Petri nets and behavioral profiles,based on behavioral profiles,an indirect inheritance matrix is established based on the indirect dependencies of event logs.A large number of event logs are divided into different regions.Then,based on the coupling between different regions,process models are extracted by using process mining algorithms.(2)In terms of process model modeling and optimization,this paper divides the process model into different regions,according to a certain region when not satisfying customer requirements,process optimization for a certain region,according to the indirect dependency of event log tasks to establish the process sub-model,And continue to optimize the model,and then interact with the sub-model business process model.This article verifies the feasibility of the method through online shopping system.
Keywords/Search Tags:Process mining, Petri net, behavior profile, event log, indirect inheritance, model optimization
PDF Full Text Request
Related items