Glossary · Semantic interoperability and information models
Meta-model
Also known as: Metamodel
German: Metamodell
In modeling, a meta-model is a model that defines the elements, relationships and rules that can be used to build models in a particular modeling language or framework. For example, the AAS metamodel defines what an AAS, a submodel and a property are, and the UML metamodel defines classes and associations.
- Information models
In one sentence
A meta-model defines the elements, relationships and rules from which models in a given modeling language or framework are built.
Example
The OPC UA address space model is a meta-model: it defines node classes and references, which companion specifications use to model machines.
How it applies
- Engineering: Meta-models underpin modeling tools and exchange formats. Tools that share a meta-model can exchange models more reliably.
- Integration: When mapping between standards, such as AAS and OPC UA, the meta-models have to be mapped first, before the actual data models.
- Documentation: For readers, clearly separate the levels: the meta-model (what kinds of elements exist), the model (a specific machine type) and the instance data (a specific machine). Mixing these levels is a common source of confusion.
Meta-model vs. model
A model describes a system or domain using a language; a meta-model describes the language itself. A class model of a valve is a model; the UML definition of what a class is belongs to the meta-model. A Schema (data) plays a similar role for data formats.