Research And Implementation On Changing Mining Of Process

Process management system is the use of computer technology to automate implementation of the organization in some or all of the business processes, which can greatly increase the efficiency of business processes and implementation of quality. However the mechanism has led to the serious differences between the prescribed process models and the real running process instance. Modeling process need not only to business modeling officers familiar with the operation of the process and also have the ability to process modeling techniques, which means the course of the differences between the real work of the process and the process by the management system provided is inevitable. The Exposure of these problems make the process management mechanism has not only failed to make business process management automatic or semi-automatic completion of the business, but let this become a mechanism for the implementation of the business of a burden rather than help. Therefore, in order to enhance the effectiveness of the process management system must address the modeling process and the actual process of bias, the bias is usually attributed to the change of process schema.On the basis of analysis of the origin and development of the process mining technology, thesis summarized the current research on it and pointing out a number of shortcomings and the coming work in the near future. Using the concepts and methods of the process mining, the changing mining by change log from the process management system of some process schema has been discussed,whose aim is to extract the change information of the original process schema, in order to support modify, extend or redesign the process.Mining Change information is a knowledge-discovery process, the paper focused on two stages: the change log pretreatment and changes mining. Based on the form of change organization-change transaction, the relationship between primary changing operations and its concomitant changing operations has being discussed. According to this, a purging algorithm of process change logs with complex changing transactions is proposed, which can deal with the irrelevant information and noise. Then experiments show the correctness and efficiency of the method. Afterwards, the thesis discussed the dependence and the commutativity of change operations. And based on multi-phase process mining algorithm, a two-step change mining algorithm with the dependence and the commutativity of changes has been proposed. We derived the correctness as well as efficiency only by formal proof of the correctness of the algorithm, but by using platform ProM to the develop a two-step change mining algorithm plug-in, with MXML change logs, to give experimental change mining. And the change process as the result of change mining has been analyzed and explained, which the verified the algorithm. Finally, this article summing up our work, and proposed some of the prospects of change mining and related studies for future work and more in-depth research direction.