Optimization Analysis Of Business Process Behavior Consistency Based On Process Mining

Many companies need to adopt efficient and effective ways to analyze their behavioral consistency,which can improve the efficiency of business operations to adapt to the current rapidly changing market.Due to the various needs of consumers and the current open environment,the precise acquisition of reference models is often not a breeze and it is more often in the form of logs,so there are defects to detect the relevant properties in this case.Therefore,it is of great theoretical and practical means to explore how to use process mining techniques to detect the existence of anomalies quickly and accurately.Most of the existing consistency analysis methods presuppose the existence of reference model and the absence of data change,and then the behavior deviation of the source model and the reference model is analyzed by using behavioral profiles and other measures.Due to the large number of business process structure in many current network systems and the fact that the data will change to a certain extent due to the interference of various factors during operations.If the existing technology is continuously used to detect abnormalities and find behavioral deviation areas,it will consciously ignore the error information resulting from the data changes,which will lead to reduced accuracy of the results.Based on the above background,in order to improve the efficiency and accuracy of the process consistency analysis with no reference model existence and the change of data,and to quickly find out and optimize the inconsistent regions,this article from the following points of view to conduct an in-depth investigation:(1)Regarding the problem of mining business process models with complex structures and in open environments,the existing research is only limited to closed systems with relatively simple structures.This paper takes the Open Petri net as the premise and by using the communication behavioral profiles between interactions to construct model,and then applies this method to a specific mobile phone recharging system instance to verifies its feasibility.This method not only reduces the time of the complex network model,but also helps to further detect whether the behavior is biased and repair the behavior deviation region.(2)The existing technology focuses only on the perspective of control flow and rarely considers the influence of data information,which has certain deficiencies in detecting the deviation of the behavior under the changing data.In this paper,we raise a similar optimal alignment consistency analysis means for detecting data-constrained models,which not only fully takes into account the impact of data information on the presence or absence of behavioral deviations,but also compensates for the defaults of the existing ways that only result in static analysis of data change,and overcomes the deficiencies of the reduced accuracy of results,which only measure the consistency of the control flow and ignore the impact of data change propagation.(3)In order to solve the problem of repairing the behavior deviation region,this paper uses the concepts of behavior inclusion,after pinpointing the behavior deviation region,then searches for all model segments in the repository that can replay deviant behavior in the log and its behavior that can contain model deviation regions behavior.Then the closeness score is used to determine the optimal matching segment set,and then connects the replaced fragment with the completed model to obtain an optimized complete model.Finally,the optimal repair results are selected by comprehensively considering the structural similarity and consistency between the source model and the repaired model,so as to realize the effectively and efficiently repair the behavioral deviation region to achieve a more realistic simulation of the actual operation.