An analysis of OWL-based semantic mediation approaches to enhance manufacturing service capability models

The exchange of accurate, computer-interpretable information is critical in today’s dynamic supply chains in which manufacturers come and go as needed. This exchange begins when manufacturers, who hope to join the supply chain, provide the OEM (original equipment manufacturer) with information regarding their production capabilities. These capabilities are represented electronically in what are called manufacturing service capability (MSC) models. These models are frequently proprietary, which makes them difficult to access, and imprecise, which makes them difficult to use. Web Ontology Language (OWL) is a powerful language for capturing the semantics of such models. OWL can enhance both precision and accessibility, but it requires semantic mediation to resolve semantic conflicts and more importantly to enhance model semantics. Semantic-mediation approaches can generally be classified into two approaches mapping-based and reference-ontology-based. This paper characterises and compares the two approaches. Characterisation is based on examples of proprietary MSC models and by deployment criteria including mediation quality, scalability, evolution, and knowledge organisation. Comparison is based on the behaviours and trade-offs of the two approaches in the context of these deployment criteria. The paper also provides a decision-making template associated with these criteria. Finally, the paper uses this template to show under what conditions each mediation technique is preferred.