A Practical Approach to Ontology-Based Data Modelling for Semantic Interoperability

Abstract

Efforts to provide a standard representational framework based on current materials modelling and characterization knowledge facilitating collaboration, digital data representation, knowledge systems and semantic interoperability has been the main agenda for the European Materials Modelling Council (EMMC). One challenge in adopting the technology is related to the on-boarding process, particularly regarding the availability of ontologies and practical aspects to linking data to the ontologies, which requires deep insights in the current knowledge organization system. Latter challenges we address by constructing an interoperability framework based on data-models and explicit ontological mappings. Data models allow for building a representation of a computer system from different perspectives. The conceptual view identifies what real-world concepts the data represents. The logical perspective defines the rules and structures of how to implement the strategy. The physical perspective establishes the relationship between the data model and a specific database system. We show the ontology-based data modelling approach. It addresses the separation of the logical and the conceptual perspective, and allows for the development of the data-models. This approach also puts the data-models directly into production before applying the ontological mappings or developing the domain ontologies. Finally, we show a framework for information exchange that connects data models to physical storage and allows software applications to eliminate the need to support specific input-output operations, file conversion, file versioning, etc.

Full document