Authors:
Nadejda Alkhaldi
1
;
Sven Casteleyn
2
and
Frederik Gailly
3
Affiliations:
1
Vrije Universiteit Brussel, Belgium
;
2
Universitat Jaume I and Vrije Universiteit Brussel, Spain
;
3
Ghent University, Belgium
Keyword(s):
Enterprise Modelling, Enterprise Ontology, Ontology-driven Modelling, BPMN.
Related
Ontology
Subjects/Areas/Topics:
Artificial Intelligence
;
Data Engineering
;
Enterprise Information Systems
;
Information Systems Analysis and Specification
;
Knowledge Engineering and Ontology Development
;
Knowledge-Based Systems
;
Ontologies and the Semantic Web
;
Ontology Engineering
;
Symbolic Systems
Abstract:
Within an enterprise, different models – even of the same type - are typically created by different modellers. Those models use different terminology, are based on different semantics and are thus hard to integrate. A possible solution is using an enterprise-specific ontology as a reference during model creation. This allows basing all the models created within one enterprise upon a shared semantic repository, mitigating the need for model integration and promoting interoperability. The challenge here is that the enterprise-specific ontology can be very extensive, making it hard for the modeller to select the appropriate ontology concepts to associate with model elements. In this paper we focus on process modelling, and develop a method that uses four different matching mechanisms to suggest the most relevant enterprise-specific ontology concepts to the modeller while he is creating the model. The first two utilize string and semantic matching techniques (i.e., synonyms) to compare t
he BPMN construct’s label with enterprise-specific ontology concepts. The other two exploit the formally defined grounding of the enterprise ontology in a core ontology to make suggestions, based on the BPMN construct type and the relative position (in the model). We show how our method leads to semantically annotated process models, and demonstrate it using an ontology in the financial domain.
(More)