Datenbestand vom 20. Januar 2019
Tel: 089 / 66060798
Mo - Fr, 9 - 12 Uhr
Fax: 089 / 66060799
aktualisiert am 20. Januar 2019
978-3-8439-0404-9, Reihe Informatik
Consistency Checking for Ontology-Based Workflows
202 Seiten, Dissertation Technische Universität Kaiserslautern (2011), Softcover, A5
Workflow Management Systems are of increasing importance in research and industry. Consistent workflow descriptions are essential to exploit the positive effects of these systems, since inconsistencies usually lead to unexpected behavior during execution of the workflow, resulting in potentially huge cost and work efforts. Static consistency checking techniques promise to put an end to inconsistent workflow descriptions or at least to reduce inconsistencies significantly.
Developing such techniques is challenging, because workflow descriptions can be complex and integrate several different perspectives as e.g. control flow and data. In the last years several consistency checking techniques for workflows have been developed to check the control flow, but few approaches consider the data perspective.
In this thesis, we consider ontology-based workflows, which are used to collect and integrate complex data. In such workflows, semantics of data are described in terms of a given domain ontology. In order to guarantee reliable workflow executions, collected data has to be consistent with respect to the domain ontology and it is crucial that routing decisions based on data do not lead to defects, as e.g. workflow stalls. In such scenarios the data perspective becomes central. Hence, there is a need for data consistency checking techniques, which can detect and eliminate data inconsistencies at design time. In order to enable such techniques, a formally defined data perspective is needed.
In this direction, this thesis motivates and describes a workflow language for ontology-based workflows and explains its semantics, especially focusing on a formally defined ontology-based data perspective. Based on the semantics we define possible data inconsistencies. We present techniques for detecting such inconsistencies utilizing Semantic Web reasoning and prove soundness and completeness of the techniques.
Starting from this theoretical work, we present the architecture and implementation of the OWoCoCheck system, a flexible software system to check ontology-based workflows for data inconsistencies. We illustrate practical relevance of our approach, by outlining the possible integration of the OWoCoCheck software into the ontology-based clinical trial management system ObTiMA which stores workflows according to our workflow language.