Data interoperability in research
Data interoperability makes it easier to exchange, combine, and reuse research data seamlessly across different platforms and tools, regardless of the size of a dataset. This is especially important when working with diverse and complex datasets, or when conducting interdisciplinary research. In these contexts, data must remain understandable and reusable across disciplines with different conventions, standards, and terminology.
Benefits of making data interoperable
Interoperability removes barriers to collaboration, ensuring that data can be shared, accessed, and reused with minimal friction. This allows researchers from different disciplines to integrate information from disparate sources, supporting collaboration, actively driving innovation, and opening the door to new questions, methods, and discoveries that would be difficult to reach in isolation.
Data interoperability levels
Interoperability can be considered at several levels. Each level addresses a different potential barrier to exchanging and using research data.
Technical and syntactic interoperability
Technical and syntactic interoperability
Technical interoperability allows different systems, software, hardware, and platforms to connect and exchange data. Syntactic interoperability ensures that the exchanged data follows a structure that the receiving system can understand. This is the most basic yet critical interoperability level.
To achieve this, data must be stored and encoded in common machine-readable formats like CSV, XML, or JSON. These formats act as a universal language, allowing systems to read and process the data without compatibility issues. However, simply choosing a common file format is not enough. The structure of the data also needs to be clear and consistent.
For example, if one system records dates as DD/MM/YYYY while another uses MM/DD/YYYY, the same data may be interpreted differently. Similarly, differences in field names or data structures can make it difficult to combine datasets.
When preparing your data, consider:
-
Using common, machine-readable file formats.
-
Using consistent structures and field names.
-
Documenting how variables and values are represented.
-
Clearly defining formats for dates, numbers, and other data types.
Semantic interoperability
Semantic interoperability
Semantic interoperability ensures that research data retains the same meaning when it is used across different systems, datasets, and disciplines. In this second interoperability level controlled vocabularies, standard terminologies, and consistent naming conventions act as a shared dictionary to prevent misunderstandings.
Without semantic alignment, combining data from different sources can be misleading, as identical terms can carry different meanings in different domains. For example, field can refer to a region influenced by a force in physics, a plot of land in agriculture, or a set of numbers in mathematics. Even measurements can cause confusion: a length of 10 inches in one dataset might appear as 25.4 centimetres in another, making the data unusable without clear unit definitions.
When preparing your data, consider:
-
Using established terminology and controlled vocabularies where available.
-
Clearly defining variables and concepts.
-
Recording units of measurement.
-
Providing sufficient metadata to explain the meaning and context of your data.
Organizational and legal interoperability
Organizational and legal interoperability
Organizational and legal interoperability ensures that data can be shareable and reusable across institutions while adhering to laws, regulations, and governance frameworks. Clear agreements on data ownership, licensing, and access rights are essential, especially in fields where data may be sensitive or restricted. Without these agreements, even well-structured and meaningful data can be trapped in bureaucratic limbo.
When planning to share or reuse data, consider:
-
Who owns or controls the data.
-
Which licences or terms of use apply.
-
Who is allowed to access and reuse the data.
-
Whether personal or sensitive data requires additional safeguards.
-
Whether agreements between collaborating organizations are needed.
Planning early for data interoperability
Incorporating data interoperability at the initial design phase of your project is essential. This approach helps:
-
Reduce technical, organizational, and semantic barriers later on.
-
Enhance long-term usability, enabling seamless reuse and integration of data.
-
Avoid costly redesigns, saving time and resources in later project stages or workflows.
Frequently Asked Questions (FAQs)
Is interoperability only important for large or complex datasets?
Is interoperability only important for large or complex datasets?
No. Even small datasets benefit from interoperability. When data is combined or reused, differences in formats, structures, terminology, or units can lead to errors and additional work, regardless of the size of the dataset. Interoperability ensures your data is usable by others (and your future self) from day one.
Is using CSV or JSON formats enough to make my data interoperable?
Is using CSV or JSON formats enough to make my data interoperable?
No. While CSV and JSON are widely used for data exchange, they don’t guarantee interoperability. These formats ensure syntactic compatibility (i.e., the data can be read), but true interoperability requires semantic alignment: shared definitions, units, and metadata. Without this, data may be technically shareable but practically unusable.
How should I consider interoperability at the project design stage?
How should I consider interoperability at the project design stage?
By integrating interoperability into your research design, you can avoid costly technical, semantic, organizational, or legal barriers later. Addressing interoperability after the project is finished is both time-consuming and expensive. Decisions about formats, data structures, terminology, metadata, and access conditions are easier and more effective to make before data collection and processing begin.
The concepts described on this page are extracted from the following publication:
Mazaheri, M., Zwiers, K., Baccinelli, W., Avadakkam, S., Schreurs, E., Verhein, J., & van Gorkum, I. (2025). An introduction to interoperability in the research context. Zenodo. https://doi.org/10.5281/zenodo.17566839