Time Prediction Method And Application For Business Process Based On Queue Mining

Process mining is a supplementary for Business Process Management.Starting at event logs,it is a technology which extracts information from those logs and then discovering,monitoring,anglicizing and optimizing the actual business process.At the same time,modeling technology is usually combined with it to make the analysis results more intuitive.Process mining plays an important role in many aspects of Business Process Management,such as improving enterprise productivity,conducting the operation of enterprise and saving operating costs.Therefore,in recent years,it has been a hot topic of research.With the constant development of science and technology,process mining technology is also facing challenges.Because of the continuous improvement of the information system that makes the running data about business process can completely be found and exploited,the current process mining technology,based on control-flow and data-flow,merely aim to analyze the running state of the process without time delays have shown some limitations.It is shown especially in the circumstance of process running that a large number of multi-class tasks which need to be executed at the same time,that is,the execution of the tasks in a business process are delayed.To solve this problem,this paper proposes a method of business process optimization based on queue mining.First of all,starting from the event log,the initial model of the business process is developed by using the theory of behavior profile combined with incremental log;Secondly,by using the idea of queue mining,the customer's behavior(delay)information can be found out through the time delay prediction of specific customers,so as to achieve the purpose of optimizing the process model;Finally,an example is given to demonstrate the effectiveness of the proposed optimization method.The optimized process model not only has a good replay effect to the event log,but also can reflect the behavior information of the task that with multi-class and time delays.The other main contributions of this paper are as follows:(1)Present a method of time delay prediction for multi-class queue mining.On the basis of queue mining,combined queuing theory,this paper put forward different delay prediction algorithm according to G/M/s+M,D/M/c+M and M/M/1 three different type of queue system.In addition,in consideration of the effect of different customer categories in the service process,predict the time delay based on the delay of the queue length for target-customer at the instant.On the one hand,this method expands the application range of the queue mining technology,on the other hand improves the current queue mining method in the consideration of multi queue type.(2)Present a method of service resource allocation based on queue mining.Resource allocation scheme in the service process means that how to select the resources to provide services to customers.Reasonable resource allocation scheme plays an important role in guiding the service operation and ensuring the quality of service.But the current research on the resource allocation scheme has some disadvantages in considering the interaction between resources and customer.In this paper,we propose a service resource allocation method based on queue mining.Firstly,according to the field constraint,resource matching is carried out on the basis of the attributes of resources and customers to ensure the fairness of the service process.Secondly,based on the queuing perspective and considering the interaction between resources and customer,the queue information of customers and the decision information of resources in the service process are achieved,and then the scheme is optimized by using that information to improve the efficiency of allocation.Finally,the effectiveness of the proposed method is verified by an example analysis and simulation experiment.(3)Present an optimization model of service interaction process model based on configuration information.The service interaction process model is interacted of two aspects of the process model,which is composed of the customer process model and the resource process model,and optimizing the service interaction process model has a good effect on improving the resource utilization and customer service experience.At present,for the service process research mainly concentrated in the resource scheduling and customer behavior analysis,however,there are limitations in the study of on the impact of resource scheduling on customer behavior.In this paper,a method of optimizing the service interaction model based on configuration is proposed.Firstly,the behavior profile of customer events in service flow is proposed,and start with the service log in the service process,the behavior information is obtained by using the weak order relation of the behavior profile;Then,two new configurations are proposed based on the queue mining,which is based on the service process model and the resource allocation scheme,And according to the customer behavior information,using the configuration structure of the service interaction process model to analyze the fitness and to detect the consistency of behavior,through the optimization analysis to obtain the final model of service interaction process.Finally,the effectiveness of the proposed method is verified by an example analysis and simulation experiment.