基于案例属性的子目标挖掘方法

Subgoals Discover Method Based on Case Attribute

为了解决从日志中挖掘出更精确的过程模型问题,提出了一种基于案例属性和时间相关性的挖掘子目标的方法。首先利用案例属性对原始案例进行分组,得到具有特定属性的案例子组;然后通过案例目标与子目标间的时间相关性,挖掘每个子组中的子目标。最后基于挖掘出的子目标对案例应用过程发现算法,挖掘出的模型可以对基于原始案例挖掘的模型进行优化和改进。通过在ProM上对真实案例进行实验表明,利用基于案例属性的子目标发现方法,得到的模型比基于完整日志发现的模型具有更好的质量。

In order to solve the problem of mining more accurate process models from logs,a method of mining subgoals based on case attributes and time correlation is proposed.First,use case attributes to group the original cases to obtain case subgroups with specific attributes,then use the time correlation between the case goals and the subgoals to mine the subgoals in each subgroup.Finally,based on the excavated subgoals,the case application process discovery algorithm is used,and the excavated model can optimize and improve the model based on the original case excavation.Experiments on real cases on ProM show that by using the subgoals discovery method based on case attributes,the model obtained has better quality than the model discovered based on the complete log.