Compliance Monitoring of Data-Aware Declarative Process Models

Abstract

Runtime Compliance Monitoring in Business Process Management is an area that is dedicated to the early detection of non-conformance between a set of business constraints and the events that are recorded by an Information System supporting the execution of a business process. Companies need to make sure that their business practices conform to predefined requirements so that, in case of non-conformance, it is possible to take appropriate action to mitigate any further damage. Runtime monitoring makes it possible to continuously check the state of a process execution w.r.t to a set of business constraints. This thesis focuses on a Runtime Compliance Monitoring technique that is able to check along with constraints over the control flow of a process also constraints over the data perspective of a business process. For modeling business constraints Multi- Perspective Declare (MP-Declare) is used. Alloy model checker is used for checking individual constraints at runtime as well as possible conflicts between two or more business constraints. The approach developed in this thesis was evaluated with synthetic and real life event logs and it was also compared against another existing approach. This implementation is also integrated as part of a process mining application.