The Scientific Variables Ontology (SVO) is a blueprint outlining all of the required and optional components of creating a machine-interpretable scientific variable concept. SVO comprises an upper ontology (SVU) that contains a set of modular conceptual categories and defines required and optional relationships between them, as well as a lower ontology (SVL) that contains instances within each category. SVL can be manually or automatically populated with concepts from a wide range of fields to create custom variables.
SVO originated in the idea of creating standard names for model integration. Standard names are by necessity reliant on a controlled vocabulary to ensure that any given term or set of terms has a one-to-one mapping to a specific meaning. An ontology, on the other hand, is concept-based and thus terminology agnostic. In order to allow seamless mapping from ontology to standard name, each granular concept in the ontology is labeled with a
skos:prefLabel that, in combination with a set of rules for assembling a standard name, can be used to automatically generate standard names. Due to its customizable nature, it is not necessary to be restricted to one set of rules or one vocabulary for generating standard names. Rather, it is possible to introduce a custom serializer along with custom preferred labels for different domains, and thus generate standard names following different rules, as necessitated by a given application.