A multi-commodity flow based approach to scheduling multi-model synchronization

Process models are used widely in business process management related initiatives, such as business process optimization, IT system implementation, etc. In large amount of practices, multiple process models are maintained simultaneously and are required to keep consistent with each other. Synchronization of multiple process modules has been studied as a combination of several two-process-model synchronization problems for a long time. However, it introduces a new problem that has few research results: scheduling the synchronization to make all modifications in all models propagated to their targets in minimized time or steps. This paper proposes an approach to the scheduling problem based on multi-commodity flow network problem while taking several special constraints into consideration. The process models and the available synchronization channels are considered as the nodes and arcs in the network respectively. Through optimizing the flow of model modifications in the network, the synchronization of multiple process modules is scheduled aiming at improving synchronization efficiency and ensuring model consistence.