Welcome to the Scientific Variables Ontology website! We are happy you are here. Below are a few pointers to help get you started!
SVO is a framework that enables modular, principled expression, manipulation, and creation of scientific variable concepts. It provides:
Scientific variables are present in all kinds of digital resources. From being mentioned within unstructured web or journal articles, to being explicitly recorded in structured digital data resources and being transformed or computed by scientific models, variables are at the heart of how we communicate about scientific observations. Since variables are not identified in a standard way across resources, it is difficult to identify matched variables for the purpose of automating the creation of scientific workflows and the generation of new knowledge from available information. Although there are currently techniques for detecting (likely) similar concepts using natural language processing (NLP) techniques, these techniques can only approximate meaning, and results from such techniques may miss critical links or may create false connections between concepts. SVO aims to provide both a concept-based blueprint for representing variables in a machine-readable format as well as ontology-guided tools (APIs) for generating new variable concepts from freeform text and matching freeform text to existing variable concepts.
The foundational concepts of the ontology—Phenomenon, Process, and Property—are illustrated in the image (credits) above. These concepts commonly occur in a variety of available upper ontologies. The scope of SVO is to provide the rules and modular constructs necessary for the custom creation of scientific variable concepts, from an observer's perspective.
This project is supported by funding from DARPA and NSF EarthCube.