Query data inconsistency for business processes

Business processes are designed to achieve business goals under procedural rules by orchestrating tasks, information and documents. Managing data inconsistency in business processes is a challenging task. If not managed properly, business will face negative financial consequences. For instance, in a typical service business process, dispatching invoice depends on service being ready. When starting to dispatch invoice, if the service is not yet deployed, the invoice will be wrong and there may be financial damage to the business. From literatures, some work provides modelling for business process execution with notations of activities, events and flows and deals with inconsistency problem by patterns; some work analyses data generated in business process and investigates the reachability between data points. Although substantial works have been done, the data inconsistency problem has not been properly resolved. In particular, it is still lacking of modelling language and resolution for inconsistency caused by multiple starting points of business processes and dynamics of business processes execution. This thesis provides data consistency solution in three aspects: a business process modelling in enriched business workflow notation with data states and temporal properties, a classification of data consistency categories, and a workflow query algorithm to discovery data inconsistency issue.