Workflow Log Process Mining And Abnormal Diagnosis Based On Petri Net

Process mining of business processes has always been a core issue in the field of business process management.With the advent of the e-commerce era,major companies and organizations pay more and more attention to business processes.How to effectively dig out useful information from a huge business process system is critical to the development of companies and organizations.important.Process mining technology is a technology that uses the event logs generated during the daily operation of the business process system to find a more reasonable process model,and combines it with other methods including consistency detection,change domain analysis and other methods to optimize the business process system and model.The final model.The research of process mining technology can help reduce the operating costs of enterprises and organizations,enhance the property of commodities and services,and at the same time enable enterprises and organizations to better adapt to social and economic development.At present,there are many process mining methods for business processes,but many of them have disadvantages such as large amount of calculation and low degree of automation.In response to the above problems,combined with the mathematical knowledge learned,I think of rough set theory that has a good effect on dealing with uncertain problems.Using rough Petri nets to re-examine the process of mining problems and find that it has good results in some cases.It not only simplifies the process,but also gives a formal mapping,which lays the foundation for future automation.In the process of mining some useful information,you will inevitably encounter various problems.How to locate the existing problems quickly and accurately is very important for the business process system.Process mining is based on the event log generated by the business process system,so most In this case,the problems encountered in process mining are also caused by the event log.Through learning and summarizing,I understand that several common problems encountered in process mining are related to the abnormal types of event logs,so I want to establish a mapping between the two through the mapping matrix in mathematics to achieve the production in process mining Anomalies,as long as you perform a simple operation on the mapping matrix,you can understand the cause and effect of the anomaly based on the results.Everyone knows the computer's ability to process numbers.If the computer's powerful computing power can be used to realize the process mining of business processes,its efficiency can naturally be significantly improved.How to convert the event log into a digital matrix that the computer can process is to achieve this.One of the most primary manners of the measure,after consulting a large number of documents,it is understood that the causal relationship matrix has this transformation ability.The main investigation contents of this paper are as follows:(1)Aiming at the process mining methods of existing business process workflows,which generally have a large amount of calculation and are not fully automated,a rough Petri net-based workflow process mining method is proposed.It makes use of the advantages of rough set in dealing with uncertain problems.First,it converts the event log generated during the operation of the business process system into a rough Petri net,which only reflects the most basic description of the generated event log,and then according to the rough The set rules transform the rough Petri net into an intermediate process model,and finally according to the proposed algorithm,the intermediate process model is transformed into the final Petri net model,so that it can match the event log.(2)In view of the uncertainty,complexity and ambiguity in the relationship between abnormal signs and abnormal causes in workflow log process mining,a method for diagnosing abnormality in workflow log process mining based on rough set theory and Petri net is proposed..First introduced the basic concepts of Petri nets,and then introduced the related concepts of rough set theory.Through the matrix algorithm of discernibility matrix algorithm and rule extraction in rough set theory,the conditional attribute reduction and rule simplification of the decision table are realized,and the decision table is found Hidden potential rules,reduce workload,reduce the impact of uncertain information,and combined with the ability of Petri nets to parallel inference,realize efficient process mining abnormal diagnosis through simple matrix operations on workflow logs,and finally an example analysis proves this Effectiveness of the method.(3)Due to the complexity of expression of the model and event log,it is difficult to directly input the computer for automatic analysis.In order to solve this problem,it must be transformed into a form that can be processed by the computer,so a method of mutual transformation between the process model and the causal relationship matrix is proposed,and the algorithm is used to process the causal relationship matrix converted from the original model,and finally The processed causality matrix is re-transformed into a Petri net model,and various indicators are compared based on the given event log with the original business process model,and a model that is more consistent with the given event log is selected.At the same time,because it can be converted into a matrix form,it also lays the foundation for using computers to deal with such problems in the future.Figure[26]table[5]reference[94]...