Translating Disciplinary Knowledge: Model Coupling through Ontologies
Abstract
In a multi-disciplinary environment a common understanding of concepts and their relationships is needed for successful cooperation between disciplines. To achieve a common understanding between models - that is a model provides inputs to other models in a coherent way - first the modellers should understand and translate the knowledge that they let their models to exchange. The aim of this paper is to illustrate the potential usefulness of knowledge bases and ontologies in making knowledge explicit and re-usable between different models, exchanging data with spatio-temporal, biophysical and economic dimensions. We will present a case study based on the SEAMLESS project, which applies ontologies to a set of economic models, based on different methodologies, e.g. empirical econometric estimation models versus a mechanistic optimization model operating across different scales and one biophysical model, e.g. a dynamic crop growth simulation model. An ontology in computer science is considered as a specification of a conceptualization. After several iterations during our collaborative approach in which a number of scientist participated, a common ontology was developed. Within this common ontology the ontologies of the individual models can be distinguished, just as the links between these ontologies through shared concepts. We thus demonstrated how models can be linked through meaningful inputs and outputs, which are stored as concepts in an ontology. It is concluded that ontologies help to rigorously link models of different structures from different disciplines in a meaningful way, and an ontology can be beneficial in further ensuring that scientific knowledge is salient, legitimate and credible.
Sign up today - FREE
Mendeley saves you time finding and organizing research. Learn more
- All your research in one place
- Add and import papers easily
- Access it anywhere, anytime