Abstract: Most of the existing conformance detection methods are based on the individual activities alignment,and only a kind of fast matching method between model and logs is given,which deducing a problem that the non-block structure model is insensitive.A process conformance detection method based on the alignment of direct succession relationship was proposed.The related properties of model and logs were obtained through the direct succession relationship,and the relevant attributes in the log were displayed in the way of adjacent activity pairs.According to the optimal alignment algorithm based on the adjacent activity pair,the optimal matching activity pair sequence of trajectories was obtained.Then the fitting degree function of the single trace and the model based on the minimum cost was proposed,and the fitting degree algorithm was presented,thus the fitting degree of the logs and the model was calculated.Practical cases were used to evaluate the proposed method.The results showed that the proposed method could correctly calculate the fitting degree between process model and log,and had better space-time complexity.

Key words: process mining, conformance checking, adjacent activity pairs, alignment

摘要： 针对现有的符合性检测方法大都基于单个活动的对齐,且只给出模型与日志的快速匹配方法,存在非块结构模型不敏感的问题,提出一种基于直接后继关系对齐的过程符合性检测方法。首先,根据直接后继关系得到模型与日志相关属性,并以紧邻活动对的方式展现出来。其次,根据基于紧邻活动对的最优对齐算法(AAP),得到轨迹的最优匹配活动对序列,提出基于最小代价的单条迹与模型的拟合度函数,并给出拟合度算法(DFA),计算日志与模型的拟合度。最后使用实际案例对所提方法进行了评估,结果表明：该方法能够正确计算过程模型与日志之间的拟合度,且具有较好的时空复杂度。

关键词: 过程挖掘, 符合性检测, 紧邻活动对, 对齐