Cookies policy

GNOSS usa cookies, propias y de terceros, con finalidad principalmente técnica y necesaria para prestación de nuestros servicios y mostrarles contenido relevante. Más información en nuestra política de cookies.

ACEPTA para confirmar que has leído la información y aceptado su instalación. Puedes modificar la configuración de tu navegador.

CASOS DE ÉXITO

Where art finds its digital voice

Starting point

For decades, museums have maintained their core resources — art collection, historical archive and specialist library — as entirely separate silos, each with its own cataloguing systems, documentary languages and access logic. This fragmentation, inherited from distinct disciplinary traditions and incompatible technological systems, forces researchers, curators and visitors to consult each resource independently, losing valuable connections and making it harder to develop a comprehensive understanding of the heritage.

The Museo de Bellas Artes de Bilbao identifies this separation as a fundamental limitation that impoverishes the experience of knowledge. A researcher studying a Basque artist must physically consult the archive to locate correspondence or historical photographs, visit the library to access specialist bibliography, and explore the collection to identify their works — with no way of visualising these connections in an integrated way. This fragmentation not only consumes time and resources, but prevents the discovery of hidden relationships between documents, works and bibliographic references that could open up new interpretations.

The museum proposes a paradigm shift under the motto "Where art finds its digital voice". This is not about digitising isolated resources or creating a simple online catalogue, but about establishing a genuine dialogue in which artificial intelligence amplifies the human capacity to discover, understand and create meaning from cultural heritage. The museum's website must be able to communicate with everyone, offering each user what they need to know at any given moment, advising and recommending what is most relevant — transforming fragmented consultation into an experience of continuous discovery.

Solution overview

A landmark in semantic integration

The Museo de Bellas Artes de Bilbao sets an international precedent as the first museum to complete a full semantic integration of its three core pillars — art collection, historical archive and specialist library — into a single Knowledge Graph queryable by both machines and people.

What makes this truly disruptive is not the digitisation of each resource in isolation, a practice already widespread in museums around the world, but the creation of a system that understands the relationships between all of them. This digital knowledge representation system understands facts relating to artists, works of art, content, subjects, periods and styles, as well as any object linked to them — and crucially, it understands how all these entities are connected. The platform fundamentally transforms the experience of discovering and understanding the museum's heritage, offering possibilities for exploration and contextualisation that go far beyond the constraints of physical space and time inherent in consulting these resources in the traditional way.

This integration resolves the longstanding problem of fragmentation through a technical architecture that overcomes barriers which have remained insurmountable for other museums: the incompatibility between documentary standards across different disciplines, the difficulty of establishing meaningful semantic relationships between heterogeneous resources, and the complexity of maintaining intellectual coherence whilst integrating tens of thousands of records from different systems and cataloguing traditions.

Where art meets intelligence

The Museo de Bellas Artes de Bilbao sets an international precedent as the first museum to complete a full semantic integration of its three core pillars — art collection, historical archive and specialist library — into a single Knowledge Graph queryable by both machines and people.

What makes this truly disruptive is not the digitisation of each resource in isolation, a practice already widespread in museums around the world, but the creation of a system that understands the relationships between all of them. This digital knowledge representation system understands facts relating to artists, works of art, content, subjects, periods and styles, as well as any object linked to them — and crucially, it understands how all these entities are connected. The platform fundamentally transforms the experience of discovering and understanding the museum's heritage, offering possibilities for exploration and contextualisation that go far beyond the constraints of physical space and time inherent in consulting these resources in the traditional way.

This integration resolves the longstanding problem of fragmentation through a technical architecture that overcomes barriers which have remained insurmountable for other museums: the incompatibility between documentary standards across different disciplines, the difficulty of establishing meaningful semantic relationships between heterogeneous resources, and the complexity of maintaining intellectual coherence whilst integrating tens of thousands of records from different systems and cataloguing traditions.

The result is a dialogue that unfolds across multiple levels:

  • For the curious visitor, the system acts as an expert guide that reveals unexpected connections, suggests personalised itineraries and contextualises each work across multiple layers of meaning. Someone exploring a Basque landscape painting will automatically discover historical photographs of the same location in the archive, articles about the region's industrialisation in the library, and works by other artists who documented the same territorial transformation.
  • For the researcher, it offers an unprecedented tool in which 70,475 archive documents engage in dialogue with 26,698 works of art and 61,578 bibliographic resources, revealing patterns and relationships that were previously invisible. What once required days of physical consultation across different museum departments now comes together in a single search, presenting all related information in a unified and contextualised way.
  • For the educator, it provides pedagogical resources that can be adapted to different levels and educational contexts, generating personalised thematic itineraries in real time according to the specific needs of each group.
  • For the curator, it brings together in real time all available documentation on each work — from preparatory sketches in the archive to documented restorations and technical bibliography — facilitating discoveries and new lines of research that had remained hidden by the separation between resources.

Where art finds its digital voice

Knowledge at scale: the graph by numbers

The Museo de Bellas Artes de Bilbao has consolidated a graph comprising 10,187,378 entities and 36,216,679 relationships (RDF triples) — a remarkable knowledge density that enables deep connections across every element of the collection. This semantic architecture manages:

  • Archive documents: 70,475
  • Bibliographic resources: 61,578
  • Works of art: 26,698
  • Authors/artists: 2,218
  • Exhibitions: 740
  • Multimedia resources: 3,420
  • News items: 505
  • Storage: 1 TB of digitised images and documents
  • Relational density: 280 triples per work

The integration of 70,475 documents from the historical archive represents a genuinely distinctive advantage. Whilst other major museums have prioritised enriching their collections with external references — links to Wikipedia, international databases or third-party resources — the Museo de Bellas Artes de Bilbao has taken a different strategic path: deeply integrating its own documentary holdings to create a highly dense, cohesive graph specialised in Basque art and its context. This approach generates connections rooted in original, exclusive knowledge that belongs to the museum alone, and is difficult for other institutions to replicate.

BM ontology network: the architecture of knowledge

The project implements what we call the BM Ontology Network — a complex semantic architecture that enables the system to understand not only what each element of the museum is, but how it relates to all the others within a web of cultural meanings. This technical complexity is necessary because the collection, archive and library follow radically different documentary logics, developed over decades by distinct professional communities with their own incompatible standards.

The ontological model of the Bilbao Museoa / Museo de Bellas Artes de Bilbao hybridises three fundamental international standards, each specialised in its own domain, enabling them to communicate coherently for the first time:

  • CIDOC-CRM (ISO 21127:2014) as the central core — the international standard for cultural heritage documentation developed by the ICOM International Committee for Documentation. CIDOC-CRM provides the descriptions and formal structure to represent the museum's core processes, from the cataloguing of works to conservation processes and educational activities.
  • ICA RiC-O (Records in Contexts Ontology) for the management of the historical archive, developed by the International Council on Archives. This ontology enables precise modelling of the 70,475 archive documents, their contexts of creation, the relationships between documents, and their connection to the works and artists in the collection.
  • FRBRer (Functional Requirements for Bibliographic Records) for the library — the IFLA entity-relationship model that structures the 61,578 bibliographic resources and enables the representation of complex relationships between works, expressions, manifestations and items within the museum's documentary holdings.

This hybridisation is complemented by ontologies for enrichment and interoperability, including DBpedia for links to Wikipedia and encyclopaedic context, SKOS for thesauri and controlled vocabularies, and Schema.org for web metadata.

Semantic bridges between documentary worlds

The power of this architecture lies in its ability to build semantic bridges between documentary worlds that have traditionally never communicated. The ontological network enables the system to understand, for example, that a sketch in the archive (managed with RiC-O) is related to a finished work in the collection (modelled with CIDOC-CRM), that both are linked to a monograph in the library (structured with FRBRer), and that all of this connects to an exhibition documented in historical photographs — creating a complete, contextualised narrative that no museum professional could reconstruct manually without investing days of work.

The model follows W3C Semantic Web standards (RDF/OWL) and the Linked Data Principles, ensuring the highest quality in terms of data openness, portability and interoperability. This complex ontological architecture consolidates the museum's full body of knowledge assets into a single digital ecosystem, enabling for the first time a true semantic integration of collection, archive and library.

Impact

This pioneering integration brings to life the concept of "intelligent dialogue" at the heart of the project. A search for any Basque artist does not simply retrieve their works in the collection — it also surfaces archive documents that mention them (correspondence, historical photographs, exhibition records) alongside all related bibliographic material from the library, presenting everything in a single, unified and contextualised record, in real time. This capacity for automatic synthesis transforms queries that once required multiple visits to different museum departments into immediate moments of discovery.

A superior knowledge experience

The platform incorporates seven transformative capabilities that turn the promise of "where art finds its digital voice" into concrete functionality:

  • Intelligent semantic metasearch: Rather than matching words, it understands concepts and intentions. A search for "Basque industrial landscape" recognises that the user is not looking for those three words in sequence, but is interested in a complex cultural phenomenon. The system retrieves works depicting industrialisation, archive documents about factories and the territorial transformation of Bizkaia, bibliography on the Bilbao estuary, and connections between artists who documented this transformation — presenting an integrated narrative where previously only scattered fragments existed.
  • Faceted navigation: Faceted search emulates human reasoning, allowing users to refine their exploration by combining criteria intuitively — period, technique, subject, artist — whilst discovering unexpected connections suggested by the system based on the semantic relationships within the graph.
  • Auto-generated contexts: Each work is automatically presented enriched with all related information available in the museum: linked archive documents, specialist library bibliography, works connected by artist, subject, iconography, period or technique, and stories and multimedia content. This enrichment, which previously required manual work by documentalists for each specific case, is now generated automatically across all 26,698 works in the collection.
  • Contextual recommendation systems: The system understands the semantic content of the museum's assets to suggest personalised exploration itineraries. If a user looks up a work by Regoyos, the system does not simply show other works by the same artist — it proposes a journey that may include the artist's correspondence in the archive, monographs on his technique in the library, works by contemporaries who explored similar themes, and historical context about the period in which he lived.
  • Semantic dynamic publishing: The museum's writing and documentation processes are automatically consolidated into the knowledge graph, enriching and extending it without duplicating information across different systems. The system can automatically generate specialist publications — virtual thematic exhibitions drawn from any subset of the knowledge graph, such as "Basque art between the wars", "women artists in the collection" or "representations of industrial labour" — multiplying the possibilities for communicating and disseminating the heritage without adding to the museum team's workload.

Ultimately, the Museo de Bellas Artes de Bilbao project demonstrates that when advanced technology is placed at the service of culture with rigour and conceptual sensitivity, it does not replace the human experience of enjoying, contemplating and understanding art — it radically enriches it, contextualises it in ways previously impossible, and makes it accessible to everyone, at any time and from anywhere, whilst always maintaining the excellence our cultural heritage deserves. The intelligent dialogue between user and heritage ceases to be a metaphor and becomes a technical reality that transforms the way we discover, study and enjoy art.

Built with

Graph Discovery AI capabilities
Semantic AI Platform Platforms
Visual AI capabilities
Context AI capabilities
Ontology AI capabilities