Discovering Inconsistencies Caused by Inappropriate Data Operations Based on Workflow Nets with Tables

Business processes without inconsistent data errors resulting from inappropriate data operations are regarded as a fundamental and important subject in major application areas. At present, such errors are always explored by only focusing on process data or persistent data; studies have seldom considered both data layers. In fact, each business process contains the data of such two layers, and its execution depends on the interaction between them. This paper employs WFT-nets (workflow nets with tables) to model business processes, considering process and persistent data. It extends the WFT-net's reachability graph to capture process data and persistent data interactions, proposes an automated inconsistency detection method, and introduces a data refining strategy to enhance detection accuracy and workflow repair efficiency. Based on WFT-nets, inappropriate data operations that can lead to the error of inconsistent data are given in formula form. The extended reachability graph displaying the data operations of the WFT-net is constructed under the corresponding data refinement. Then, inconsistent data are detected by traversing such reachability graph. Some theoretical arguments are given; a set of comparative experiments with WFD-nets were conducted to illustrate the effectiveness of our inconsistent data detection approach.