Business process model is a reflection of the computerized model, is under an advanced computer environment to be able to support the implementation of business process integration and business process automation and the set up by the Business process management system to implement the business model. The life cycle of the workflow consists of four phases:workflow design, workflow configuration, workflow execution and workflow diagnosis. The process mining is not just a business process workflow design tool, but it is very helpful for understanding the current business process. The aim of process mining is to reverse the process and collect data of performance to support workflow design and analysis.This paper firstly introduced the latest state of development of the mining areas of the business process, then introduced the current reference model and modeling method, and then describes some current difficulties in the mining areas, the introduction of this paper is to study and solve a difficult point:Non-free choice structure.Before discussing the mining method, some technology and theory related to process mining are described, including algebra express of log and definition and property of Petri Net and Workflow Net, creating the mapping relationship between Petri Net and Workflow Net. Then paper describes the current process mining field of relatively perfect a-algorithm and heuristic algorithm, pointing out shortcomings and limitations of them on the mining of some structures.In this paper, due to the deficiencies and drawback of the existing process mining algorithm, such as the structure of non-free choice and so on. The current main process mining algorithms can not discover the structure of it, mainly because they can not dig out the non-local non-free-choice construct. Before introducing the proposed method, we define:the tasks, the judge rules of non-free choice. Tasks defined in the paper determine the fundamental basis for the rules of mining algorithms, to express the order between the tasks in the log; the judge rules can infer the relationship of the non-local non-free choice between some tasks according to the addition between the tasks. They make a theoretical preparation for later proposed the process mining algorithms.Then we propose a new process mining algorithm that can discover the structure of non-free choice, which is based on the flow of the event log. Events recorded in the log is analyzed, and then divided into tasks at all levels. Through the judge rules, we can infer the non-local non-free choice relationship between the task pair. The algorithm can also dig out the choice, parallel, and mixed structure of process models.Finally, we implement a process mining algorithm plug-in of the Prom buy using java, and get business process diagrams by digging a series of business process instances. At the same time by generating and running a number of experiment al data, the mining quality of the algorithm is analyzed, and it is proved that th e algorithm has obvious advantages n terms of complex structures, such as non-f ree choice, it affords an effective solution to the deficiencies and limitations of t he a-algorithm in mining non-free choice structure. |