A Framework Proposal for Intelligent Management of Unexpected Exceptions in Workflow

Workflow management systems are recognized to provide essential and effective support to business process management. However, many events, either too infrequent or outside the business descriptions, are generally not taken into account and consequently, are not integrated in the a priori definition of the workflow. This kind of exceptions are known as unexpected exceptions. Usually, handling them relies on human interventions that are potentially error prone and could take a lot of time. In this work, we propose a solution to manage unexpected exceptions. To that end, we aim at providing a framework that supports real time automatic management of these exceptions. It relies on the definition of a reaction to an exception that is guaranteed to be optimal and adapted to the workflow context. In particular, we use artificial intelligence techniques to generate the reactions automatically