Research On Business Process Implementation And Mining Technology For Vehicle Sales System

With the rapid development of Internet,the process of enterprise information is getting faster and faster,it is gradually transformated to service industry which focuses on customers and business.Sales service is an important part of car service.Vehicle sales business process system plays an important role as a collaboration platform within enterprise or between enterprises.This thesis mainly solves the following problems through the analysis of business process model for vehicle sales business and the status of sales system.1.Enterprise of sales service areas have individual needs,but the current business process systems are not perfect in the personalized customization for business process.2.Asymmetric information between business analysts and system developers may cause the problem that developers may not fully meet the needs of business people.3.Business processes are encoded in the fixed system function,and it is difficult to be adjusted according to the actual needs.In view of these situations above,BPMN is selected as the standard language of the process modeling scheme by comparing the existing business process modeling methods in this thesis.The online process designer with graphical interface may perform simple process customization.In order to achieve the frequent process changes of process system,the vehicle sales business system based on process customization and process secondary development of Activiti Process Engine is proposed.This system can take out the business process logic form software code and realize the separate management of business processes,which can have a positive response to customized mutative processes.Taking the vehicle sales business process as an example,the functions of process customization,uploading,deploying,circulation and data query are displayed.Finally,the process mining algorithm based on instance log is proposed which reappears the business process structure of instance by using the measurement method of C-D Vector Similarity,K-means clustering algorithm and dependency calculation.