The project tackled a critical gap: the lack of a cohesive framework for linking media art information with consistent metadata, terminologies, and identifiers. 'Connecting Media Art' improved usability by linking to standardised terms through the DDHN Terminology Network, publishing the heritage information as linked data, and making a media art dataset findable to others through the DDHN Dataset Registry. The project set out to address a pressing question: how can media art collections become more discoverable, interoperable and sustainably accessible across institutions and borders? Rather than creating yet another platform, the focus was placed on the infrastructure beneath the surface: the data, standards and shared agreements that allow collections to connect across institutions, systems and borders. The project has delivered a solid foundation for long-term, international collaboration; and it is only the beginning. The first building blocks are now firmly in place, creating a platform for future collaboration and development across the media art field.

The approach was intentionally iterative. To maintain focus and deliver tangible results, the project started small with one core concept (media artwork type) and one clearly defined subset of the collection: single-channel video works. This choice enabled rapid testing, validation and refinement. Future phases will expand this groundwork towards more complex media art forms, including multi-component installations and software-based works.

Key Results

The project delivers a set of practical, immediately applicable building blocks:

1. Sustainable identifiers: A robust URI strategy is realised within the standards under the mediakunst.net domain, enabling sustainable, institution-independent identification of artworks, artists, expressions, manifestations, items and terms.

• Unique URIs within the network domain, for example mediakunst.net: https://id.mediakunst.net/w-00010107,
• Simple and recognisable, with a prefix per entity (for example w = work, a = artist, e = expression, m = manifestation, i = item, t = term),
• Applicable to both the LI-MA collection and to partners connected via mediakunst.net.

2. Standardised terminology: The project adopted a phased approach to standardised terminology. A first step towards a shared media art thesaurus was taken, starting with media artwork types and aligned with the Termennetwerk; artists are linked via Wikidata QIDs, ensuring interoperability and consistent identification. This improves data connectivity, reuse, exchange and collaboration while reducing future corrections.

3. Linked Open Data publication: To enable meaningful connections between collections, data is published as Linked Open Data using GraphQL endpoints. This allows both people and machines to access, search and link collection data, helping users discover relationships between artists and artworks. The endpoints, including a visual editor, available through https://graphql.mediakunst.net/, support multiple formats, including JSON(-LD) and Turtle. As a result, the data can be reused across different systems and workflows, forming part of a broader network of cultural heritage information.

4. Registration in the DDHN Dataset Register: The LI-MA and Van Abbemuseum datasets of single-channel video artworks have been registered in the DDHN Dataset Register, making them discoverable within the national digital heritage infrastructure. This dataset was selected because it represents a significant part of the collection and provides an effective basis for testing and refining the data model before extending it to more complex media art. Registration includes metadata, a URI, institutional information and licensing details, improving accessibility and reuse. Check the links:

• https://datasetregister.netwerkdigitaalerfgoed.nl/dataset?uri=https://data.mediakunst.net/dataset/lima-single-channel-artworks
• https://datasetregister.netwerkdigitaalerfgoed.nl/dataset?uri=https://data.mediakunst.net/dataset/van-abbe-single-channel-artworks

5. A clear data model: the Four Level Artwork (FLA) Model: The project developed the Four Level Artwork (FLA) Model, based on FRBR (Functional Requirements for Bibliographic Records entity-relationship model), to reflect the layered nature of media art. The model distinguishes four levels: Work, Expression, Manifestation and Item (WEMI). This enables complex relationships between artworks, versions and presentations to be documented more accurately than a simple one-record approach. Tested on single-channel video artworks, the model bridges media art collection practice and international linked data standards and provides a foundation for future expansion.

Four Level Artwork (FLA) Model, 4 levels:

• Work: Conceptual top layer: 'the concept' or 'the underlying idea' of a work of media art,
• Expression: Concrete, conceptual level: the idea applied, a concretisation of a conceptual idea,

Manifestation: Concrete technical realisation of an Expression (recursive): the materialisation of the Expression (the carriers, configurations, or builds),

• Item: Physical or digital representation of the Manifestation (a specific tape, file, server copy, etc.).

6. Working in an DDHN-compatible way and the role of the CIS: The project made LI-MA’s Collection Information System DDHN-compatible, positioning it as both a data source and a link to DDHN services. Collection data (as well for partner collections) is published as linked data, uses shared terminology, and is registered in the Dataset Register. This improves findability and reuse while keeping data at the source. The approach reflects the DDHN principle of 'register once, reuse widely' and shows how a collection system can function as a node in a wider data network.

7. Case study Van Abbemuseum: The Van Abbemuseum media art collection was used as a case study to test the project approach in practice. Within a subcollection of single-channel video artworks, identifiers, terminology and the new data model were applied, while existing metadata was migrated, harmonised and validated. This produced a complete workflow from source data to connected dataset. The resulting insights were translated into practical guidelines and good practices for use by other affiliated institutions.

8. Knowledge sharing: The project’s findings were shared through expert meetings, workshops and publications. Topics included practical steps for collection holders, such as identifiers, terminology, Linked Open Data, dataset registration and data modelling, as well as collaborative approaches to defining shared standards and achieving interoperability. Beyond developing technical infrastructure, Connecting Media Art also strengthened a community of practitioners.

Together, these results demonstrate that interoperability is not an abstract ambition but an achievable, practical reality, provided institutions align on shared principles and standards.

From Pilot to Practice

The project’s central insight is straightforward and actionable: start small, think big. By focusing on manageable subsets and clearly defined steps, identifiers, terminology, data publication, registration, institutions can build momentum without losing direction. The expansion path is already clear: from single-channel video to installations, from controlled vocabularies to richer semantic networks, from isolated datasets to a fully connected knowledge graph.

Crucially, this is not a closed exercise. The work has been developed in continuous dialogue with the field, nationally and internationally, and will continue to evolve in that same collaborative spirit. Follow-up projects are already in preparation, extending the model, deepening the datasets and addressing challenges such as rights management, IIIF integration and complex digital artworks.

A Shared Infrastructure in the Making

'Connecting Media Art' does not claim to be finished. It marks the point where fragmentation begins to give way to structure. Where isolated collections start to function as a network. Where media art gains the digital infrastructure it requires, and deserves.

For collection holders, from museums and archives to festivals, galleries and independent artists, the project offers concrete tools to position their collections within a broader ecosystem. For researchers, professionals and the wider public, it opens up new ways of discovering, understanding and connecting media art across contexts.

'Connecting Media Art' may have reached the end of its first chapter, but the larger story of connected media art collections is only just beginning.

Publication

To mark the completion of the project, LI-MA is publishing the Dutch-language public report: "Connecting Media Art: How Linked Open Data and Collaboration Connect Media Art Collections". The report presents the project's results, lessons learned and future ambitions in an accessible format for collection holders, heritage professionals, researchers, artists and the wider public. An English-language edition of the report will be published shortly.

This report sets out not just what has been achieved, but how to proceed. The message is direct: the groundwork is in place. Now we build, together.

For more information or to get involved, contact LI-MA via info@li-ma.nl.

Header: Ulay and Marina Abramovic, AAA-AAA (1978).