The usefulness of process models (e.g., for analysis, improvement, or execution) strongly depends on their ability to describe reality. Conformance checking is a technique to validate how good a given process model describes recorded executions of the actual process. Recently, artifacts have been proposed as a paradigm to capture dynamic, and inter-organizational processes in a more natural way. Artifact-centric processes drop several restrictions and assumptions of classical processes, e.g., process instances cannot be considered in isolation as instances in artifact-centric processes may overlap and interact with each other. This significantly complicates conformance checking; the entanglement of different instances complicates the quantification and diagnosis of misalignments. This paper is the first paper to address this problem. We show how conformance checking of artifact-centric processes can be decomposed into a set of smaller problems that can be analyzed using conventional techniques.
|Number of pages||39|
|Publication status||Published - 2011|