Research On Configuration Of Business Process Model And Log Repair Based On Behavior Deviation

Due to the variability of business process model in practical application and the load of log system,the behavior of event log and business process model are often inconsistent,that is,behavior deviation.However,the analysis of behavior deviation can effectively improve the quality of the model.In order to improve the fitting degree of event log and business process model,the model configuration analysis and log repair are good methods.Model configuration analysis technology can not only set different configuration types for activities according to the special needs of users,but also optimize and simplify the model to achieve full replay log while preserving the value of the original model.Log repair is one of the new technologies in business process management(such as cloud computing and big data).In view of incomplete logs,relevant methods are used to repair the deviation between logs and models,which not only retains the value of original logs and models,but also maintains the similarity between them to the maximum extent.Therefore,model configuration and log repair are extremely important.According to the behavior relation between Petri net and log activity,this paper compares the behavior relation between the given initial process model and event log,to find the behavior deviation,analyze the cause of the problem,and achieve the optimization effect according to the method of model configuration analysis and log repair.The main contents of this paper include:(1)Regarding the behavior deviation between business process model and event log,existing researches mainly focus on the analysis of configuration based on personal theory experience and the separation of change and configuration.There are still deficiencies in how to correctly feedback configuration information and correctly configure.Therefore,based on the existing model configuration algorithm,propose an algorithm to extract behavior deviation and implement model configuration based on behavior relation for logs containing behavior deviation.Based on the behavior similarity,the behavior deviation is detected and the location of the deviation is located,and the configuration information is fed back by the logical propositional algorithm to effectively implement the model configuration.(2)For the problem of predictive alignment of business process models,most of the existing alignment literature realizes optimal alignment by comparing and analyzing process model processes with brute force search method,ignoring the efficiency of alignment.Therefore,in order to break fixed pattern,realize prediction,and make process management more efficient and accurate,this paper proposes an alignment prediction method based on sequence pattern.In this method,conditional probability is used to effectively select the activities to be aligned,and activity partitioning technology is used to ensure the efficiency and accuracy of alignment.(3)Further research and analysis of log repair problem based on model constraint.As for the problem of log repair,the existing researches are mainly based on trace alignment technology to repair incomplete log,however,there are still deficiencies in the efficiency and accuracy of repair.Therefore,this paper proposes a repair method for logs.Under model constraints,the method uses conditional probability and confidence level to obtain a complete optimal repair result,and utilizes model partition technology to effectively improve the repair efficiency of the repair method.Figure [18] Table [19] Reference [85]...