An introduction to interoperability in the research context

1 An introduction to interoperability in the research context A collaboration initiative between interoperability community managers Maryam Mazaheri Maastricht University Library Lena Karvovskaya TDCC-NES Kimberley Zwiers Health-RI/TDCC-LSH Walter Baccinelli Health-RI/TDCC-LSH Sreenithya Avadakkam Vrije Universiteit Amsterdam Emma Schreurs University of Amsterdam Jessica Verheijen Tilburg University Ingrid van Gorkum DCC-PO 2 Contents Introduction: Why Interoperability Matters in Research............................................... 3 The Role of Standards in Achieving Interoperability ..................................................... 3 Levels of Interoperability .......................................................................................... 4 Barriers and Misunderstandings in Practice ............................................................... 5 Strategies for Enabling Collaboration ........................................................................ 5 Difference between Interoperability and Reusability (in data contexts) ........................ 6 References .............................................................................................................. 8 3 Introduction: Why Interoperability Matters in Research Today's research landscape is more collaborative and data-driven than ever before. In a university setting, where disciplines range from computational science to humanities, researchers generate diverse types of data and software using a variety of tools and platforms. In such a setting, interoperability is the foundation for meaningful collaboration. Interoperability is the ability of different systems, datasets, and platforms to work together by seamlessly exchanging and interpreting data. It ensures that data can be shared, accessed, and reused without significant technical or organizational barriers. Interoperable data is machine-readable, understandable, and reusable without losing its meaning. In research, interoperability is vital for fostering collaboration within and across disciplines, enabling innovation through the efficient exchange and use of data. However, achieving this is challenging due to technical, structural, and organizational obstacles (Mistry et al., 2022). Despite growing awareness, these obstacles are often only fully recognized during later phases of research projects, when misalignment becomes more difficult and costly to address. Embedding interoperability considerations early in project design can significantly reduce friction and improve longterm data usability. The Role of Standards in Achieving Interoperability A major barrier to interoperability is the lack of standard practices. Adopting standards is essential for enabling consistent, accurate, and efficient data exchange. Standards relevant to data interoperability operate at several levels, such as technical, syntactic, semantic, organizational, and legal. By embedding these standards into everyday research practice, institutions can transform fragmented datasets into shared resources, accelerating interdisciplinary discovery, improving research reproducibility, and amplifying the societal impact of their work. Practical applications of interoperability in this context include structured data flows between institutional repositories, research information systems, and domainspecific archives. When implemented effectively, these systems not only ensure compliance with open science policies but also enhance the visibility and reuse of research outputs across disciplinary boundaries. In the next section, different levels of interoperability standards will be presented. 4 Levels of Interoperability Technical and Syntactic Interoperability At the most fundamental level are technical and syntactic standards. Technical standards relate to system connectivity, such as compatible networking protocols, hardware interface matching, and communication protocol compatibility. These ensure that systems can physically connect and exchange data. Syntactic standards govern the structure and format of data, enabling systems to parse and process it correctly. For example, using a shared, machine-readable format such as CSV, XML, or JSON allows data to move between systems without relying on specific software. Syntactic interoperability is a critical first step, without which systems cannot exchange information effectively. However, syntactic alignment alone does not guarantee meaningful integration. Even when using common formats such as CSV or JSON, inconsistencies in field naming, date formats, or structural assumptions can lead to parsing errors or misinterpretation. Adherence to shared schemas and validation tools is necessary to ensure syntactic interoperability across systems. Semantic Interoperability Semantic standards ensure that the meaning of data is interpreted consistently across systems. For example, the term current can refer to the flow of water in an oceanography dataset, the flow of electricity in a physics dataset, or the present time in social science research. Similarly, temperature values could be recorded as 25 (°C) in one source and 77 (°F) in another without clear unit metadata. Without semantic alignment, merged datasets risk serious misinterpretation. Shared controlled vocabularies, standardized units, and consistent metadata are essential to preserve intended meaning during exchange. Semantic interoperability requires not only the adoption of standard vocabularies, but also consistency in how those vocabularies are applied. Domainspecific ontologies and modelling tools, such as those used in healthcare or cultural heritage, are increasingly employed to define unambiguous relationships and improve cross-disciplinary understanding. This is particularly important in interdisciplinary data environments, where identical terms may carry divergent meanings. Organizational and Legal Interoperability Organizational, legal, and social standards also contribute to interoperability. Organizational policies, data-sharing agreements, intellectual property rules, and privacy regulations determine whether and how data can be exchanged. Addressing these factors is just as important as resolving technical, syntactical, and semantic issues. Legal and organizational alignment – supported by trusted relationships and shared understanding across teams and institutions – plays a critical role in creating an environment where data can be responsibly reused. Clarity around licensing, ownership, and data access permissions is often a prerequisite for cross-institutional collaboration, especially when sensitive or restricted datasets are involved (Awada, Phillips & Bogdan, 2022). 5 Horizontal and Vertical Interoperability Interoperability is not just a static feature of systems or standards—it is about how data and tools interact across institutions and throughout the research lifecycle. ● Horizontal interoperability refers to compatibility between systems or datasets at the same stage of research, such as repositories across institutions using the same metadata profiles. It enables cross-disciplinary or cross-institutional collaboration and data reuse. ● Vertical interoperability ensures smooth transitions between different stages of the research process—from planning and data collection to analysis, publication, and preservation. It allows tools and systems at different abstraction levels to interact seamlessly, reducing manual work and preserving data integrity throughout the lifecycle. Vertical interoperability is enabled by workflow systems, APIs, and containerization technologies that connect data pipelines across stages. Without it, research becomes fragmented and error-prone, limiting reproducibility and scalability. Barriers and Misunderstandings in Practice Despite growing institutional support for interoperability, several common misunderstandings persist in research practice: ● The assumption that using common file formats alone (e.g., CSV or JSON) ensures interoperability, when structural and semantic alignment are still lacking. ● The tendency to postpone interoperability planning until the final stages of a project, often leading to costly retrofitting of metadata and formats. ● The belief that interoperability only applies to large-scale collaborations, rather than to smaller datasets that could serve future analysis or reuse. ● The assumption that adopting standards is sufficient, without ensuring shared interpretation and consistent implementation. ● The view that interoperability is static, rather than a dynamic, evolving process influenced by user needs, tools, and context. Strategies for Enabling Collaboration Achieving effective interoperability depends not only on standards and infrastructure, but also on successful collaboration between researchers and data stewards. Key enablers include: ● Co-designing data workflows and governance frameworks from the outset, ensuring relevance and usability. 6 ● Providing discipline-specific metadata templates and vocabularies to reduce researcher workload. ● Aligning support timelines with the research lifecycle. ● Recognizing and rewarding data stewardship contributions within institutional frameworks. Interoperability is Contextual and Evolving Interoperability is not a binary state.t exists on a spectrum, depending on the data, users, context, and tools involved. The same dataset might be highly interoperable for one audience, yet unusable for another without additional metadata, context, or structural changes. Key considerations include: ● There are always two parties: Interoperability is not just a feature of the data. It is a communication bridge between provider and user, with both having different expectations and needs. ● Context matters: What seems “obvious” to one discipline (e.g., temperature unit, project background) may be entirely opaque to another. Without explicitly encoding context, especially in interdisciplinary work, semantic drift or misinterpretation becomes likely. ● Interoperability degrades over time: As standards evolve, users change, and technologies update, even once-interoperable data may become difficult to reuse. Planning for long-term interoperability means embedding flexibility and documentation. These challenges highlight the need to explicitly define for whom and for what purpose the data should be interoperable and to revisit those assumptions over time. Difference between Interoperability and Reusability (in data contexts) In FAIR contexts, Interoperability and Reusability have a significant apparent overlap for non-specialists, which can confuse researchers and data stewards. The key difference (for the research output provider) is that ● Reusability information is mainly intended for human users, describing how the data or software should be used (and when not) in research contexts. Examples include readme-files and other human interpretable documentation. ● The Interoperability information is more related to automated systems and structures, describing in a very formal way the content and context, potential usage, relevant standards it matches, and interoperable interfaces. Examples 7 include specific metadata describing the contents and context of the data using specified (linked) ontology, linkages to other research products, lists of (linked) standards fulfilled etc. In the time of Large Language Models, the difference will be blurred, but the more formal the description is, including encodings of context and usage, the more relevant and reliable results LLMs and other automated tools can be expected to produce. Both information types can be used as a source for Findability information, enabling usabilitybased searches. 8 References Awada, L., Phillips, P. W. B., & Bogdan, A. M. (2022). Governance and stewardship for research data and information sharing: Issues and prospective solutions in the transdisciplinary plant phenotyping and imaging research center network. Plants, People, Planet, 4(1),84–95. https://doi.org/10.1002/ppp3.10238 Mistry, P., Maguire, D., Chikwira, L., & Lindsay, T. (2022). Interoperability is more than technology: The role of culture and leadership in joined ‑ up care. The King’s Fund. https://assets.kingsfund.org.uk/f/256914/x/c48bd5a1a2/interoperability_more_than_te chnology_2022.pdf assets.kingsfund.org.uk