Font Size: a A A

Model Mining And Optimization Based On Petri Net

Posted on:2021-05-20Degree:MasterType:Thesis
Country:ChinaCandidate:Y N LuFull Text:PDF
GTID:2428330605456945Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Business process management can not only improve the service efficiency required by the market for the enterprise,but also greatly improve the market competitiveness of the enterprise.Therefore,the mining of high quality and high efficiency business process has become the research focus of domestic and foreign scholars.In order to obtain a more complete business process model,we need to perform process mining and configuration optimization.In order to make the mining model more realistic for the business process,this paper chooses to mine the model from the system log.In addition,from the perspective of behavioral profile,the system log is analyzed and combined with Petri net theory to mine the source process model.However,due to different business process modeling purposes and different modeling roles,the models created at different levels of ion and different perspectives are quite different.At this time,the degree of consistency between the log and the model needs to be measured,and the consistency measurement method is derived.Finally,the unreasonable part of the model is configured and optimized.At this time,this step is mainly completed through three transitions in Petri net theory.The main contributions of this paper include:Firstly,Log mining based on behavior profile.This paper analyzes the weak order relation of the activity pairs in the log from the perspective of the behavior profile of Petri net,and puts forward the algorithm of mining the business process model in the form of Petri net from the event log.After selecting the log instances,the algorithm divides these instances,analyzes the transition pairs contained in the log from the perspective of the behavior profile,and obtains the correct behavior profile relation,Then the source business process model is established.Secondly,Model configuration based on consistency measures,the mining model is analyzed from the perspective of behavior profile and causal network.Configure the business process model according to the consistency measurement configuration method.The specific explanation is made in combination with the process model of hotel payment system in actual production.Finally,Model optimization based on behavior profile.Mined models will inevitably have some unreasonable structures.In order to adjust these structures,this paper proposes a model optimization analysis method,and conducts specific examples by using an intelligent express cabinet system example.Figure[18]table[14]reference[66]...
Keywords/Search Tags:petri net, model mining, event log, optimization
PDF Full Text Request
Related items