基于增强日志的过程挖掘算法

Process Mining Algorithm Based on Enhanced Log

过程挖掘的目的是通过分析系统中的日志以发现、构建和优化系统业务流程.现有的过程挖掘算法大多采用从控制流角度记录和观察进程工作流的系统日志,且日志在使用前需进行大量预处理工作将其转换为算法可识别的事件日志,不仅仅增加了挖掘难度,最终挖掘所获过程模型所含属性单一,很难准确描述实际流程的运作.为减少预处理工作,增强过程挖掘算法挖掘能力,基于系统事件执行详情,通过可利用属性的提取,提出了增强日志的概念,并基于增强日志开发出一种新的过程挖掘算法.此方法利用增强日志中的附加属性,识别事件结构,通过有色Petri网的加入,挖掘出具有场景信息的过程模型.最后通过一个具体的挖掘实例进一步说明了该方法的可行性及结果的可扩展性.

The purpose of process mining is to discover,build and optimize system business processes by analyzing the logs in the system.Most of the existing process mining algorithms use system logs that record and observe process workflows from a control flow perspective,in which way,a lot of preprocessing work needs to be done to convert it into an event log able to be recognized by the process mining algorithms before the log is used,not only increasing the difficulty of mining,but also making it difficult to accurately represent the operation of the actual process because of the single attribute of the process model.In order to reduce the preprocessing work and enhance the mining ability of process mining algorithm,based on the system event execution details,the concept of enhanced log is proposed by extracting the available attributes,and a new process mining algorithm developed based on the enhanced log.This method uses additional properties in the enhanced log to identify event structures,and with the colored Petri net added,the process model with scene information is mined.Finally,a mining example is given to further illustrate the feasibility of the method and the scalability of the results.