Font Size: a A A

Process Mining Algorithm

Posted on:2014-03-10Degree:MasterType:Thesis
Country:ChinaCandidate:X H YeFull Text:PDF
GTID:2268330401454061Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
One of the aims of process mining is to retrieve a process model from an eventlog. The discovered models can be used as objective starting points during the deploymentof process-aware information systems and as a feedbackmechanism to check prescribed models against enacted ones.Process mining is helpful for deploying new business processes as well as auditing, analyzing and improving the already enacted ones.Through to analysis the research status of process mining algorithm, Analysistheadvantages and disadvantages of process mining algorithm. According to analysis the problem of process mining field put forward improvement methods. Such as a tasks that exist in a process model but not in its event log. This is a problem since invisible tasks are mainly used for routing purpose but must not be ignored.And at the same time, the namesake task and a repetitive task exist in the business process of system logof a large number. It leads to the results of the process mining will often produceinaccurate flow model.The purpose of this study is to obtain the more accurate process model.This paper puts forward the innovativemethod to improve existing process mining algorithm on these problem.Mainly is divided into two parts.The first part, solve the problem of invisible tasks. First of all, define the log and the relationship between thelogswith the synchronization management work flow net. Thenanalysis the log.Finally, propose a new process mining method based on this,and get process model.The second part,in order to improve the accuracy of the process mining, analyzed theeponymous task and repeat task, it puts forward an improved method.The new method not only can dig thecomplex structure, such as circulation structure,free choice structure and so on, it still can dig the namesake task and repeat task in the log. The innovation of this study is that, firstly, Select synchronous work flow net as a foundation, according to analysis successfully obtain the invisible task and cycle task from logs. Secondly,according to improve algorithm make sure the accurate of the business model which obtain through process mining on the greatest degree. Thirdly, to introduce the accept of pseudo parallel and asexual reproduction, significantly improve the time efficiency of genetic algorithm.
Keywords/Search Tags:Process Mining, Business Process Management, Complex Structur
PDF Full Text Request
Related items