Ex RORU: A New Approach To Characterize The Behavioral Semantics Of Process Models

Today's information systems need to support the execution of business process and they should manage the workflow in organizations.Many organizations and enterprises with massive business processes have urgent needs for workflow management and extracting the features of process models is one of the prerequisites.The behavioral semantics of a process model are the basic features to characterize it and the ordering relations between the executions of tasks are often used to describe the behaviors of a process model.A recent paper has proposed new ordering relations with uncertainty between the executions of tasks in acyclic process models.This work also gave an algorithm to compute the relations between pairs of tasks based on complete prefix unfolding.However,this algorithm has serious limitations.It cannot work for cyclic process models and those with silent tasks and non-free-choice constructs.In this paper,we show how to overcome these problems by a refinement of the relations(i.e.,extended refined ordering relations with uncertainty,Ex RORU for short).The main contributions of our work are as follows:1.We investigate the existing characterization methods of process model behavioral semantics and analyse the pros and cons of them.We also give a detailed discussion of the basis work of our paper and point out its limitations.2.We propose the ExRORU relations with three categories(i.e.,causal,inverse causal and concurrent)and 11 types in total,which can handle all kinds of sound process models,including those with cycles,silent tasks and non-free-choice constructs.3.We design an algorithm to efficiently compute the Ex RORU matrix of a process model based on its complete prefix unfolding,which overcomes the problems of infinite executing sequences and state explosion caused by loop structure and high parallel branching constructs.Then,we discuss and prove that Ex RORU can detect behavioral differences between any pair of process models uniquely.4.We conduct experiments on real-life and synthesized models and compare Ex RORU with other mainstream methods.