How FAIR Principles and Ontologies Help Us Get More Value from Our Data

Simply collecting data isn’t enough. To truly unlock its value, data must be findable, accessible, interoperable, and reusable. In other words, FAIR. But even when data is FAIR, the way we represent its meaning plays a vital role in making systems work together intelligently. In this article, we explore how the FAIR principles and semantic interoperability through ontologies go hand-in-hand, helping organisations turn raw information into actionable knowledge.

The FAIR Principles and Their Role in Semantic Interoperability

Initially designed for scientific data management, the FAIR principles now assist various fields in making their data more valuable.

FAIR Stands for Findable, Accessible, Interoperable, and Reusable.

1. Findable

Data should be easy to locate for both humans and machines. This means using a unique identifier and having rich metadata to find the data easily.

2. Accessible

Once found, data should be easy to retrieve through standard, open protocols. Even if the data isn’t available anymore, the metadata should remain accessible. Security measures like authentication may be applied when needed.

3. Interoperable

Data should work well with other data sets and tools. This requires using common formats, languages, and vocabularies.

4. Reusable

Data should be well described and well documented with clear usage licenses so we can reuse it.

Real-World Example of a FAIR Interoperability Solution

Anna, a government officer at the Ministry of Economy, needs to check Pit’s criminal record before approving his company registration. Thanks to a FAIR approach, she can do this easily:

1. Findable

Anna wonders: “Where can I access criminal record data?”

A system tells her, “Just connect to the official source with your credentials.”

She finds the criminal records database in a government data catalog.

2. Accessible

She securely requests Pit’s record using his universal identifier.

The Ministry of Justice replies: “I trust you. Here’s Pit’s record in a standard format.”

3. Interoperable

Since the data follows a standard model, Anna’s system can understand and process it immediately.

4. Reusable

Anna attaches the record to Pit’s company registration request. No manual adjustments are needed!

At Meaningfy, we adhere to the FAIR principles. We ensure that our clients’ data is easily found, accessed, shared in a common language, and well-documented for future use.

We break down data silos, and we help our clients achieve semantic interoperability.

Credits for the picture – Open Science Training Handbook (PDF).

FAIR lays the foundation, but to build true semantic bridges between systems, we need to go one level deeper. That’s where ontologies come in. Let’s explore how structured conceptual models help make data not just interoperable, but meaningful.

Ensuring Semantic Interoperability with Ontologies

Semantic interoperability concerns the meaning of data, ensuring that all parties understand the same language.

To achieve this semantic interoperability, we need vocabularies and ontologies that define concepts, their relationships, and their usage in a structured way. This involves developing a conceptual model that represents entities within a specific domain and formalizing the relationships between them.

At Meaningfy, we follow the LOT methodology (LOT – Linked Open Terms) to build ontologies and vocabularies and adopt the best practices like those in the SEMIC Style Guide to ensure that our ontologies and vocabularies meet EU semantic interoperability standards.

Ontology Development Process

1. We collect the requirements

Before we start, we ask fundamental questions:

• What problem are we solving?
• What’s the scope of this ontology or vocabulary?
• What should it include, and what should it leave out?

We document all of this in an Ontology Requirements Specification Document (RSD), which serves as the foundation for our work.

2. We develop the conceptual model

Once we have a clear scope, we build a UML conceptual model, which acts as the Single Source of Truth (SSoT). Using our model2owl tool, we transform this model into OWL2 ontologies and SHACL shapes.

3. We publish the ontology

A vocabulary or ontology is only useful if people can access and use it. That’s why we:

• Transform the UML model (XMI) into OWL and SHACL using our model2owl tool.
• Generate OWL, SHACL, HTML, and PDF artifacts.
• Document the models in AsciiDoc for structured and accessible documentation.
• Release the model source code (e.g., GitHub).

4. Continuous Maintenance
   Ontologies and vocabularies evolve over time. We maintain them by:

• Tracking feedback and new requirements (e.g., GitHub Issues)
• Regularly address open questions and new requests
• Releasing new versions while keeping older versions accessible

 Ontology Development Challenges:

1. Outdated schemas

Sometimes, organizations still use legacy schemas that don’t align with modern standards. We review and update concepts to ensure they stay relevant.

2. Conflicting terminologies

Different teams and systems might use different words for the same thing. We resolve these conflicts by structuring concepts hierarchically and building relationships between them.

3. Finding the right experts

We rely on domain experts to provide the knowledge needed for building accurate ontologies, but finding the right people can be challenging.

4. Extracting business requirements

Understanding and capturing real business needs for the Ontology Requirements Specification Document (ORSD) requires careful discussions with stakeholders.

5. Moderating working group meetings

Keeping working group discussions on track is crucial. We ensure that meetings stay within the project’s scope and lead to concrete outcomes.

6. Reviewing ontology models is difficult

Comparing changes in UML diagrams or RDF models is not straightforward. We put extra effort into detailed reviews to ensure quality.

7. Reaching agreement on models

Getting domain experts and stakeholders to agree on how to model requirements can take time. We facilitate discussions and find balanced solutions.

8. Continuous changes

Clients will keep updating their data specification, sometimes making already developed Ontology concepts unfit for the clients needs. We solve this problem by having regular meetings with the clients and keeping our ears open for any new updates that may require changes in ontology development.

Real-World Ontology Example

In an ontology, concepts are structured hierarchically and linked through relationships. The term “Windows” can mean a house window or a computer operating system.

Without structure, systems might misinterpret the meaning. However, in an ontology, these meanings are separated into different classes, such as:

• Building Component → Window (a part of a house)
• Software System → Windows OS (a digital product)

 And this is how we achieve semantic interoperability.

 FAIR principles ensure data is accessible and reusable. But without a shared understanding of what the data means, collaboration and automation across systems fall short. Through carefully developed ontologies and structured vocabularies, organisations can achieve semantic interoperability, making their data truly useful, now and in the future. At Meaningfy, we help make this a reality every day.

Contributed by Meaningfier Jana Ahmad.

 ___

Meaningfy continues to support the European Commission’s initiatives, leading the charge toward a transparent, efficient, and interconnected European public sector. If you represent a European Institution or a public company that needs to implement an interoperability solution, visit this page or contact us for tailored support and effective implementation.

---