I-ADOPT Framework ontology

The I-ADOPT Framework is an ontology primarily designed to facilitate interoperability between existing variable description models (including ontologies, taxonomy, and structured controlled vocabularies). One of the challenges in representing semantic descriptions of variables is getting people to agree about what they mean when describing the components that define the variables. The I-ADOPT ontology addresses this by providing core components and their relations that can be applied to define machine-interpretable variable descriptions that re-use FAIR vocabulary terms. It was developed by a core group of terminology experts and users from the Research Data Alliance (RDA) InteroperAble Descriptions of Observable Property Terminology (I-ADOPT) Working Group. The first published versions of the ontology up to v0.9.1 satisfied the basic cross-domain interoperability requirements. This new version of the ontology (v1.0) includes exactly the same original components plus an extension which is designed to support the aggregation of variables using a range of user-defined coarser concepts to facilitate dataset discovery and aggregation into products.

The research community is creating or collecting ever-larger volumes of data to understand phenomena via their observable properties at various scale. Our ability to exploit these data as a common resource is hampered by a lack of interoperability in how we describe the data variable observed or measured. A large collection of independent terminology resources related to variables and tools across research domains and communities has emerged. Their complexity and diversity often overwhelm data managers and users, ironically posing barriers to data interoperability.

Great progress has already been made in providing machine-readable descriptions of sensors and their observation types through the OGC's Sensor Web Enablement SensorML, Observations and Measurements (Cox 2017) or the W3C's/OGC's Semantic Sensor Network (SSN) ontology or its lightweight ontology called SOSA (Sensor, Observation, Sample, and Actuator). However, "deep metadata" that further contextualizes observations (e.g. methodology, variables, parameters) is typically represented as coarsely qualified classes (e.g. "Procedure" or "Observed property"). What exactly falls into these classes is currently unconstrained and could be anything ranging from unstandardized free-text to standardized descriptions accessible via fully resolvable URIs. The Scientific Variables Ontology (Stoica and Peckham 2018, 2019) is one known existing principled ontological framework for decomposing and representing scientific variables in a machine-readable form.

The RDA InteroperAble Descriptions of Observable Property Terminology (I-ADOPT) WG set itself the objective to produce an Interoperability Framework, co-developed through inputs from a diverse community of terminology experts and users, for representing observable properties. This effort has a strong focus on variables observed in environmental research because it leverages existing efforts to accurately encode what was measured, observed, derived, or computed in relation to the Earth systems. But many of the principles it leans on will be relevant to or connected with other domains. The construction of the framework has been informed by a review of current practices used in the community. The working group is also iteratively testing and refining the framework through a set of in-depth use cases. Much like a generic blueprint, the refined conceptual framework will be a basis upon which terminology developers can formulate or refine their local design patterns, in alignment with others. With these, they may leverage their local resources in a collective attempt to represent complex properties observed across the environmental sciences (from marine, atmospheric, and terrestrial Earth sciences, as well as biodiversity).

For more information about I-ADOPT see the I-ADOPT pages on the RDA website. Comments on this version of the ontology (please specify the version number when commenting) can be made by raising issue tickets in this github repository.

The I-ADOPT Framework is an ontology designed to facilitate interoperability between existing variable description models (including ontologies, taxonomy, and structured controlled vocabularies). One of the challenges in representing semantic descriptions of variables is getting people to agree about what they mean when describing the components that define the variables. The I-ADOPT ontology addresses this by providing core components and their relations that can be applied to define machine-interpretable variable descriptions that re-use FAIR vocabulary terms. It was developed by a core group of terminology experts and users from the Research Data Alliance (RDA) InteroperAble Descriptions of Observable Property Terminology (I-ADOPT) Working Group. The first published versions of the ontology up to v0.9.1 satisfied the basic cross-domain interoperability requirements. It defines four classes or "concepts" (Variable, Property, Entity, Constraint), and six object properties (hasProperty, hasObjectOfInterest, hasContextObject, hasMatrix, hasConstraint, constrains). The Variable is the top concept. It represents the description of something observed or mathematically derived. It minimally consists of one entity (the ObjectOfInterest) and its Property; a Property being a type of characteristic (i.e. a quantity or a quality). More complex variables can involve additional entities, for example an entity may have the role of Matrix and/or of ContextObject(s). The framework does not capture units, instruments, methods, and geographical location information; however its usage recommendation will make explicit reference to these by connecting the I-ADOPT framework to existing and complementary ontologies.

This new version of the ontology (v1.0) adds one optional new class (VariableSet) and four optional new object properties (hasApplicableProperty, hasApplicableObjectOfInterest, hasApplicableMatrix, hasApplicableContextObject). This was necessary in order to enable flexibility in assigning optional and user-defined machine-interpretable categorizations of I-ADOPT variables under one or multiple coarser grouping concepts to facilitate dataset discovery and dataset aggregation. With the introduction of these concepts and properties, the framework enables different user communities or product developers to develop their own grouping criteria. While the Variable class must be connected to at least two classes via the mandatory properties hasProperty and hasObjectOIfInterest, the VariableSet class can have either of the new properties. Additionally, the VariableSet class can also be optionally connected to the Variable class using the property ro:hasMember from the OBO Relations Ontology.

S. J. D. Cox (2017): Ontology for observations and sampling features, with alignments to existing models, Semantic Web, vol. 8, no. 3, pp. 453–470.

Magagna, B., Moncoiffe, G., Stoica, M., Devaraju, A., Pamment, A., Schindler, S., and Huber, R.: The I-ADOPT Interoperability Framework: a proposal for FAIRer observable property descriptions, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13155, https://doi.org/10.5194/egusphere-egu21-13155, 2021.

Magagna, B., Moncoiffé, G., Devaraju, A., Stoica, M., Schindler, S., Pamment, A., & RDA I-ADOPT WG.: InteroperAble Descriptions of Observable Property Terminologies (I-ADOPT) WG Outputs and Recommendations (1.1.0). https://doi.org/10.15497/RDA00071/2, 2022.

Magagna, B., Moncoiffe, G., Devaraju, A., Stoica, M., Schindler, S. and Pamment, A. (2021) I-ADOPT Framework 1.0.0. In: RDA Virtual Plenary 17. RDA Virtual Plenary 17, 20. - 23.04.2021, virtual.

Magagna, B., Moncoiffe, G., Devaraju, A., Buttigieg, P. L., Stoica, M., and Schindler, S.: Towards an interoperability framework for observable property terminologies, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-19895, https://doi.org/10.5194/egusphere-egu2020-19895, 2020.

Magagna, B., Stocker, M. and Diepenbroek, M.: Towards Interoperability for Observed Parameters: Position Statement of an Emerging Working Group, in: Gaikwad, J., König-Ries, B., & Recknagel, F. (Eds). Proceedings of the ‘10th International Conference on Ecological Informatics: translating ecological data into knowledge and decisions in a rapidly changing world’, Jena, Germany, 24-28 September, 2018.

Stoica, M., & Peckham, S. D. (2018). An Ontology Blueprint for Constructing Qualitative and Quantitative Scientific Variables. In International Semantic Web Conference (P&D/Industry/BlueSky).

Stoica, M., & Peckham, S. (2019, September). Incorporating new concepts into the Scientific Variables Ontology. In 2019 15th International Conference on eScience (eScience) (pp. 539-540). IEEE. DOI: https://doi.org/10.1109/eScience.2019.00073

Stoica, M., & Peckham, S. D. (2019). The Scientific Variables Ontology: A blueprint for custom manual and automated creation and alignment of machine-interpretable qualitative and quantitative variable concepts. In Modeling the World's Systems Conference.