Research On Business Process Mining Based On Case Attribute

Currently,process models can be mined from event logs using process mining techniques,but process models do not fully reflect the real business processes of the case.To be able to mine a process model that better fits the event log,from the perspective of case attributes,you can compare the behavioral differences between different case sets,and you can also improve the quality of the process model,optimize the original model,and maximize the consistency with the real business process.Therefore,more and more scholars are focusing on the perspective of case attributes to conduct research on process mining techniques.On the basis of the relevant research on Petri network and process mining,this paper groups cases through case attributes,constructs process models for grouped cases,and then conducts quality assessment of process models to find problems,analyze problems,and study and control the impact of different case attributes on process models according to the way of case attribute division.The main contents of this article include:(1)Aiming at the problem that the business process in the online ticket purchase model is unreasonable.This section constructs an attribute-based Petri network ride-hailing business process model by differentiating the identity attributes of passengers.Firstly,the Petri network modeling of the ride process of passengers with different identity attributes is carried out,and then the Petri network behavior analysis method is used to optimize the model by adding the selection structure,and finally the two optimized ride process models with different identity attributes are integrated.Simulation experiments show that the fused model can meet the bounds and safety of Petri network.(2)Aiming at the problem that the goal is not strong in business process mining,a sub-target mining method based on case attributes is proposed.In order to improve the accuracy of the excavated process model,on the basis of the existing research,combined with the target refinement mode and case attributes,a method for mining case sub-targets is proposed.First,the original case is grouped by using the case attributes to obtain a subgroup of cases with specific attributes,and then the sub-targets in each subgroup are mined through the temporal correlation between the case target and the sub-targets.Finally,the process discovery algorithm is applied to the case based on the mined sub-target,and the model based on the original case mining is optimized and improved by the mined model.Finally,simulation experiments on real cases and synthetic cases on Pro M show that by using the sub-target discovery method based on case attributes,the resulting model has better suitability and accuracy than the model based on full log discovery.(3)Aiming at the problem of incomplete dependency mining in process mining,a business process dependency mining method based on case attributes is proposed.In order to improve the integrity of the dependencies in the mined process model,this paper first divides the case attributes into conditional attributes and decision attributes,and defines the dependencies and the measures of the circular dependencies,and then mines the dependency model based on the case set after the case attribute value classification,and finds different active dependency pairs by comparing the dependency matrix,and obtains a new dependency model after the model is fused,and the fused model has better quality.Figure [42] Table [10] Reference [90]...