Incremental Analysis of Temporal Constraints for Concurrent Workflow Processes With Dynamic Changes

Abstract

Nowadays, workflow process changes frequently in a fast-changing business environment. When updating a workflow process via some structural changes, one of the most important tasks is to maintain its consistency under temporal constraints. Several approaches have been developed to cope with this issue. However, they are either inaccurate in locating changed parts or inefficient in analyzing temporal constraints owing to their excessive or erroneous estimation of affected portions for the updated workflow processes. Based on a sprouting graph (a graph that records the structure and time information of all paths in a workflow process in advance), this paper proposes a novel approach to analyzing temporal constraints for workflow processes with dynamic changes. First, changed parts are located via affected collaboration execution paths in the new model, i.e., the workflow process after changes. Second, instead of updating all the elements, only necessary (changed) nodes corresponding to the changed parts are updated in the sprouting graph of the original model, i.e., the workflow process before changes. Finally, based on the updated sprouting graph, only affected temporal constraints (the temporal constraints whose partial or all paths are contained in the affected collaboration execution paths) are checked. Compared with the existing works, our approach is applicable and efficient to check temporal constraints for large-scale and complex workflow processes thanks to its much lower time and space complexity.