UMBC ebiquity
Computers & Industrial Engineering
A Framework to Canonicalize Manufacturing Service Capability Models
May 15, 2015
The ability to share precise models of suppliers’ manufacturing service capability (MSC) information is necessary to develop reliable methods that enable OEMs to efficiently configure agile and responsive supply chains. Currently, most suppliers use online tools to represent and share their MSC information in proprietary ways via proprietary MSC data models. These models have limited precision and interoperability. A semantically precise and rich reference MSC ontology can address both of these limitations and enable development of the reliable supply chain configuration methods. To effectively develop and deploy such a reference MSC ontology, semantic mediation between proprietary MSC models and the reference MSC ontology will be required. An important and challenging activity within the semantic mediation process is the mapping between a proprietary MSC data model and the reference MSC ontology. The challenge of the mapping activity is to resolve structural and semantic conflicts between the proprietary model and the reference ontology in a manner that is efficient and results in mapping structures that are simple to comprehend and maintain. This paper proposes an approach to address the challenge by preprocessing the structural representations of proprietary MSC data models for alignment with the set of modeling conventions (i.e., ontology design patterns - ODPs) that are also used in the reference MSC ontology. We call this preprocessing canonicalization. Canonicalization can circumvent 1:n, n:1, or n:m mapping statements that require complex expressions thereby simplifying the mapping activity and its resulting mapping statements. The main contribution of this paper is the design and formalization of an ODP-based canonicalization framework and its associated process in the context of description logic-based semantic mediation using the Ontology Web Language (OWL).
Article
83
Downloads: 318 downloads