Method for mining invisible tasks in business processes on basis of behavior profiles

The invention discloses a novel method for mining invisible tasks in business processes o basis of Petri network behavior profiles. The method comprises the following steps: firstly analyzing the behavior relationship of the longest log sequences with highest frequencies of occurrence; obtaining a source model according to the behavior relationship and the definition of a weak behavior profile; optimizing the source model step by step according to the sequences with default activities in the residual logs so as to find models with invisible tasks; and finally comparing the models with the source model to obtain the rationality and appropriateness of the models with invisible tasks. At present, some solutions are used for mining the invisible tasks in allusion to free-choice nets, but certain boundedness exists for the complicated process models. According to the method provided by the invention, the invisible tasks in the complicated process models can be mined and the rationality f and behavior appropriateness aB between the logs and the models. The method has relatively good promotion in the event log analysis of the business processes.