Valle de la Lengua Data Hub

A secure, federated environment for sharing, discovering and using Spanish-language data, with guarantees of data sovereignty, interoperability and traceability.

// WHAT IS IT?

What is the Valle de la Lengua Data Hub ?

The Valle de la Lengua Data Space (EDVAL) is a collaborative environment where organisations can share and use linguistic resources without relinquishing control over their information. The data is not stored in a single location: it remains where it is and is shared in accordance with the conditions set by its owners.

This model facilitates the development of language technologies, digital services and research projects related to Spanish. Furthermore, it adheres to European standards for secure data exchange, ensuring trust and protection throughout the process.

Key components

Within the ‘Key Components’ section, the federated catalogue, the San Millán Spanish Oral Corpus (COE San Millán) and the connector form the cornerstones upon which the system’s reliability and technical utility are built.

Federated catalogue

A smart directory where providers publish metadata for their resources in accordance with international standards such as DCAT-AP. The result? You can find out which language assets are available without moving or centralising the actual data, facilitating collaboration between systems, organisations and sectors, whilst maintaining control at source.

View the catalogue

Technology hub

It acts as a guardian of data sovereignty for each participant. This software functions as a secure gateway, separating permission control from the flow of information. Its purpose is to automatically enforce usage policies and licences, ensuring that data is always transmitted directly and in encrypted form.

More information

// HOW DOES IT WORK?

How does the Valle de la Lengua Data Hub work?

Unlike traditional models, where information is stored on a centralised server, the Valle de la Lengua Data Space operates in a federated and decentralised manner. This means that data is not copied to a shared cloud, but remains under the full control of the person who generates it.

// Step 1 of 6

Engagement and digital identity

Any organisation wishing to participate must first obtain its digital identity within the ecosystem.

We have implemented a centralised credential system to manage user authentication.
This system acts as a ‘digital passport’, ensuring that each participant is who they claim to be and has the necessary qualifications to operate. In this way, we eliminate the risk of identity theft.

// Step 2 of 6

Publication and discovery

Data providers (universities, cultural institutions, etc.) publish the metadata for their linguistic resources in a federated catalogue.

Observers or users can browse the catalogue to see what resources are available (for example, 5,000 hours of audio from the Oral Corpus) without accessing the actual content.
All the information is organised in accordance with international standards (DCAT-AP), so that systems can communicate with one another.

// Step 3 of 6

Definition of rules and licences

Before any exchange takes place, the “rules of the game” set out in the Rulebook must be accepted.

The provider uses technical policies (ODRL) to define how their data may be used: for research purposes only, for a limited period, or on a pay-per-use basis.
This framework ensures that the owner retains intellectual property (IP) rights and full control over their resource at all times.

// Step 4 of 6

The Connector

Each participant installs a Connector, a software component that acts as a secure gateway between the permissions control and the data itself.

When a consumer requests data, the connectors communicate with one another to verify that a valid electronic contract exists and that all usage policies are being complied with.

// Step 5 of 6

Direct transfer

Once the authorisation has been validated, the data is transferred directly from the supplier to the consumer.

The data is transmitted in encrypted form and in a sovereign manner.
The Valle de la Lengua Data Space does not store the data or interfere with its flow; it merely provides a secure channel between the parties.

// Step 6 of 6

Traceability and immutability

Although the Valle de la Lengua Data Hub does not view the content of the data, it records evidence of each exchange via the Clearing House service.

An immutable record is generated, enabling audits, resolving disputes and ensuring regulatory compliance (AI Act and GDPR).
This ensures that the use of Spanish in artificial intelligence is ethical, transparent and secure.

Step 1 Step 2 Step 3 Step 4 Step 5 Step 6

An open ecosystem for those who want to innovate with the Spanish language.

Who can take part in the Valle de la Lengua Data Hub ?

The Valle de la Lengua Data Hub is an inclusive ecosystem, open to any organisation wishing to add value or innovate using Spanish-language data, within a framework of federated collaboration.

Key stakeholders in the ecosystem

Public administrations
Academic institutions and research centres
Technology companies and developers
Cultural and heritage organisations
Other sector-specific data hubs

Membership is managed through institutional contact and prior validation.

Benefits of participating in the space

Access to top-quality assets:

Validated resources, such as the COE San Millán or specialist terminology.

Sovereignty and total control:

Each participant retains ownership of their intellectual property and decides who can access their data and how.

Legal certainty and ethics:

An operation in line with the AI Act and the European Data Act, ensuring compliance with legal and ethical standards.

Promoting internationalisation:

A solution enabling Spanish-speaking companies to adapt their content for a global audience.

Reducing innovation costs:

It enables solutions to be developed more quickly and at a lower cost.

Institutional support and public funding

EDVAL is an initiative launched by the Government of La Rioja, through the Foundation for the Transformation of La Rioja (FTR) and the National Center for Spanish Language Industries (CNIE), as part of its commitment to the New Language Economy and the development of Spanish in artificial intelligence.

The project is funded by the European Union through NextGenerationEU funds, as part of the Spanish government’s Recovery, Transformation, and Resilience Plan.

Frequently Asked Questions

What is a data space?

A data space is a collaborative ecosystem where organisations share and access data securely, in a controlled manner and in accordance with common rules.

What is a linguistic fact?

Any Spanish-language resource, whether spoken or written (text, audio), together with its metadata and models, that can be used for the development of language technologies and artificial intelligence systems.

Where is the data stored in the Valle de la Lengua Data Space?

Data is not centralised, but is generated under the owner’s control. Data exchange is decentralised and federated, travelling directly from the provider to the consumer via secure technical connectors.

What types of linguistic data can I find and share?

Spoken and written corpora, specialist terminology (health, law, finance), academic knowledge bases, language models (LLMs) and advanced speech and text processing services.

What are the eligibility criteria for taking part?

Obtain a digital identity (DIDs/VCs), accept the Rulebook (common rules), use certified connectors, and comply with European data protection and artificial intelligence regulations.

Do I need to provide any details to take part?

No. You can access the portal and the public catalogue without providing any details. However, the most common options are ‘consumer’ (to make use of existing resources) or ‘provider’ (to share assets), which do require you to provide details. You provide details depending on the role and level of involvement you choose.