Portfolios | Antal Dániel honlapja

Finno-Ugric Data Sharing Space

Tue, 31 Mar 2026 14:00:00 +0000

The Finno-Ugric Data Sharing Space is a research-driven data sharing space for reconnecting fragmented cultural heritage across languages, institutions, and countries. Developed with minimal resources as a curatorial and technical experiment, it combines semantic web technologies with participatory methods to make underrepresented Finno-Ugric cultures visible and reusable in contemporary digital infrastructures.

Rather than building a new repository, the project demonstrates how multilingual metadata, community-driven annotation, and lightweight governance models can enable cultural reconstruction and long-term interoperability in low-scale heritage ecosystems.

Udmurt Történeti Fotógyűjtemény

Tue, 31 Mar 2026 14:00:00 +0000

The Udmurt Historical Photographs Collection is a growing, open, and linked collection of dispersed historical photographs of Udmurt people and places in Udmurtia. We bring together relevant photographs (or their digital surrogates) and aim to place them under the care of the Udmurt community and researchers of Udmurt cultural history.

Our goal is to identify relevant photographs and provenance information in Estonian, Finnish, Hungarian, and other collections, and to validate, enrich, or repair this knowledge with the contribution of Udmurt communities.

For example, the reverse side of one photograph contains a difficult-to-decipher location (“Gombur-Gurt” and “Medemuka”) written in phonetic German transliteration. The image, labeled simply as “Udmurt woman,” belongs to a series in the photographic collection of the Estonian National Museum, where provenance is limited and some elements were removed in the 1950s by Soviet censorship.

Can we identify who this woman was? Where the series was taken? Who the photographer was? And can we locate missing images from the same series in other collections?

Answering these questions requires knowledge of the Udmurt language and historical cartography. Solving such cases can clarify the provenance of individual photographs and help recover dispersed or lost materials. Our collaborative work with Wikimedia Hungary and Udmurt communities demonstrates how this process can be carried out.

Short summary

Spaces of a Constructive Dialogue

Sat, 21 Mar 2026 00:00:00 +0000

This project, developed more than twenty years ago shortly after my participation in a UK future leaders programme, represents one of my earliest attempts to create a structured, collaborative platform for emerging artists and curators. Conceived and executed by a group of very young practitioners, the initiative combined curatorial experimentation, documentation, and network-building at a time when such cross-disciplinary approaches were far less formalised than today.

The archival documents reflect both the ambition and the limitations of that moment: they are exploratory, the documentation is uneven, but already articulate a clear intention to connect artistic practice with broader structures, and the selected artists and their work is timeless. Looking back, what is most striking is not the program’s polish, but its orientation toward collaboration, visibility, and long-term cultural positioning—principles that have remained central in my later work.

A decade after the project, several of the participants went on to establish themselves as significant figures in the cultural field, developing independent artistic practices and contributing to international exhibitions and curatorial platforms. For example, curator Barnabás Bencsik later became the director of the Ludwig Museum, Endre Koronczi one of the most respected and socially engaged creative artists of Hungary, the young photographer Gergő László founded Lumen, which remains an iconic cultural location in Budapest, and Samu Szemery, co-curator of the architectural interventions on the branch lines of Nógrád County railways, became one of the founders of Hungary’s Contemporary Architecture Centre. I remain grateful for their enthusiastic participation, bravery, and friendship.

For me, this work marks a formative moment: an early prototype of the ecosystem thinking that would later underpin initiatives such as Reprex and the Open Music Observatory in a much more formalised and technological form. While the documents belong to a different stage of life, they already reveal a consistent concern with how cultural value is created, documented, and sustained over time—an interest that continues to define my professional and research practice.

The programme itself, titled Spaces of a Constructive Discourse, was conceived as follows:

Spaces of a Constructive Discourse

The programme was conceived as a structured initiative to create new forms of dialogue between cultural production and public life in Hungary during a period of social and economic transition, just in the months of Hungary’s EU accession. It aimed to explore how public art and cultural interventions could contribute to addressing societal issues, particularly where traditional institutional channels were limited or ineffective. Rather than treating art as autonomous, the programme positioned it as an active medium for engaging with public concerns and social realities.

At its core, the initiative functioned as a multi-stakeholder platform, bringing together artists, sociologists, civil society actors, policymakers, and local communities. Through workshops, discussions, and pilot projects, it created a space where different forms of knowledge—artistic, social scientific, and administrative—could interact. The intention was not only to generate artistic outcomes, but to enable mutual learning and to integrate insights from cultural practice into broader policy and institutional frameworks.

A key objective of the programme was to expand participation in public discourse, particularly by including voices that had previously been excluded from decision-making processes. It sought to develop mechanisms through which cultural projects could surface local problems, articulate new perspectives, and feed these back into governance and reform processes. In this sense, the programme anticipated later approaches to participatory culture and socially engaged art, framing cultural activity as part of a wider system of societal feedback and transformation.

Operationally, the programme followed a structured workflow: identifying social problems, conducting research, implementing public interventions, and evaluating outcomes, followed by feedback loops into future actions. It also placed strong emphasis on partnerships and clearly defined roles (curators, coordinators, mediators). Taken together, it can be understood as an early attempt to design cultural production as a systemic, collaborative process—linking artistic experimentation with institutional learning and long-term social impact.

Főfotó

Fri, 13 Feb 2026 00:00:00 +0000

Fortepan is not only an open, privately run photography archive, but has effectively become a Hungarian national institution. Originally built from rescued negatives found during informal “spring-cleaning” discoveries, it has grown into a subjectively curated yet widely used visual memory of Hungary before 1989. Still operated by a civil society organisation, Fortepan today collaborates with major institutions, contributes to exhibitions and publications, and increasingly serves as a repository for private and institutional photographic collections.

I have been involved with Fortepan as a user, volunteer, and donor for many years. My most significant contribution was the localisation and recovery of the long-lost negative archive of Főfotó, a centralised photographic enterprise created after the nationalisation of private studios in socialist Hungary. This organisation documented architecture, industry, public life, and everyday scenes, but its archive disappeared during the transition of the 1990s.

Through collecting and analysing historical negatives acquired on the art market, I identified traces of this lost archive and, together with Fortepan, initiated a process that led to its partial reconstruction. At this point, no institutional buyer had the resources or the legal capacity to buy this risky find, so I stepped into as a buyer. We transferred 400,000 negatives (for comparison: more than the permanent collection of the Dutch Photography Museum) in a van to Budapest from a tiny, almost derelict village.

The 35mm rolls from the 1980s occupied the least space.

After clarifying legal and provenance issues, which took quiet some time, the 400,000 that came in large banana boxes and their original drawer structure ( the most important physical organisational system that we wanted to preserve intact) this collection lived with me, filling much of my bedroom.

After a long process, the with surviving inventory records—were recovered, digitised in part by Fortepan, and ultimately deposited in the Budapest City Archives. This process required not only historical and technical expertise, but also careful negotiation around ownership, provenance, and public accessibility.

For me, this project represents a concrete example of how photographic heritage can be reconstructed outside institutional frameworks and later reintegrated into them. It reflects a long-standing interest in the lifecycle of collections—how they are created, lost, rediscovered, and reinterpreted—and in the role that individuals and civic initiatives can play in preserving cultural memory.

WikiMuseum

Wed, 24 Sep 2025 00:00:00 +0000

Wikimuseum is an experimental curatorial concept developed with Wikimedia Eesti and Wikimedia Hungary, creating multilingual, co-curated exhibitions across Wikimedia Commons, Wikidata, and Wikipedia. It brings together dispersed cultural heritage—often never seen together physically—into structured digital exhibitions that connect institutional collections, private archives, and community knowledge.

The project explores how open platforms can support collaboration between museums, researchers, and source communities, while addressing legal, ethical, and data governance challenges.

Key exhibitions & outcomes

Livonian Traditional Dress (TextileBase integration) A cross-border exhibition connecting collections from Latvia, Estonia, and Finland, forming the first comprehensive, structured visual and data-driven presentation of the dress heritage of Liv (Livonian) people.

Mari Traditional Clothing (ERM-led multilingual exhibition) Built primarily on ERM collections and complemented by Finnish and Wikimedia Commons materials, this exhibition provides accessible interpretation in Estonian, Hungarian, English, Russian, and Meadow Mari, reconnecting dispersed heritage with both international audiences and Mari communities.

Udmurt Ethnographic Photography (ongoing, community-informed curation) A collaborative project with Wikimedia Hungary and Udmurt contributors focusing on improving provenance, correcting historical descriptions, and identifying dispersed or censored photographic materials across collections.

Key insights / lessons

Open platforms can function as curated museum environments, not only repositories, when structured with clear metadata and narrative layers
Multilingual annotation and community participation enable reconstruction of missing knowledge, especially for underrepresented or historically misrepresented groups
Digital co-curation creates new forms of collaboration between institutions, civic initiatives, and source communities, but requires careful handling of legal, ethical, and governance questions

TextileBase

Sun, 11 May 2025 15:05:10 +0100

TextileBase is a platform to collect, connect, and share knowledge about historical clothing.
It links artefacts, photographs, secondary sources, and institutional records into a multilingual, interoperable, and searchable knowledge graph.

By bringing together scattered records from museums, archives, researchers, and businesses, TextileBase enables richer stories about how people dressed, lived, and expressed themselves.

Why Textiles?

Unlike stone or metal, textiles are fragile. Few garments survive beyond a century. What we know often comes from dispersed museum pieces, old photographs, or drawings — often in different languages and formats.

TextileBase connects these fragments into one network: a living map of dress history.

Who is TextileBase for?

Museums & archives — link dispersed collections into larger platforms such as Europeana and ECCCH.
Researchers & cultural heritage projects — manage data to FAIR/8-star standards, publish reusable datasets, data papers, and visualisations.
Businesses — from sustainable fashion to cultural tourism, TextileBase supports digital product passports, provenance tracking, and heritage storytelling.

Connect artefacts, images, and texts into one collaborative, AI-supported research space.

👉 Build on shared knowledge: Technical details & offering

TextileBase is not only about preserving the past — it’s about making data usable now, for research, heritage, and innovation.

Examples from the Database

Connecting dispersed artefacts: 19th-century Latgalian garments in one linked dataset (Q142, Q180, Q179, Q181).

Connecting Artefacts Across Borders

Understanding Secondary Sources

Enriching records with secondary sources: trousers identified from a 1913 Seto photograph (Q331).

Following Shifting Place Names

Tracing shifting place names: a Livonian skirt recorded under Finnish, German, Livonian, and Latvian toponyms (Q347).

Building a Wikimuseum for Dispersed Collections

Inspired by Wikimedia Estonia’s multi-language, open-access model, we propose a virtual museum—a Wikimuseum—that brings together:

Artefacts from rural museums (e.g., Mõniste, Saatse, Värska)
Items in national museums (Estonia, Finland, Hungary)
Private collections that would never be physically exhibited together

Building a Wikimuseum: connecting dispersed collections across languages, borders, and institutions. See early prototype: Seto Traditional Clothing.

👉🏿 Preview Presentation The Concept of a WikiMuseum: WikiMuseum = GLAM Wiki + Wikibase + Data Sharing Space press F for full-screen view (to be presented at WikimediaCEE 2025 in Thessaloniki.)

Next Steps

👉🏻 Visit the TextileBase website

👉🏾 Subscribe (PDF)

👉🏼 Get in touch

SKCMDb

Tue, 19 Nov 2024 17:16:00 +0100

Please visit our poster and talk with our team members, Daniel Antal, Anna Márta Mester (librarian-data steward) and Anna Zilkova (chairperson of IAML Slovakia) on 8 July 2025 10:30–11:00 in the Gallery. You can download our poster in PDF here.

Check out the entry examples related to Albrecht’s: Missa in C (printed, Hudobné centrum) with library and webshop access points, and linked data on the composition itself Q479, linked to its recorded manifestations.

Early Data Access

Next steps

The Slovak pilot illustrates how libraries and national music centres can take a central role in a decentralised European Music Observatory that federates with the EU Open Data Portal, Europeana, ECCCH, DARIAH, and Zenodo. This poster accompanies a companion contribution on the Finno-Ugric Data Sharing Space, including the LīvMDb (Livonian Music Database), which explores how smaller regional repertoires can be embedded in broader, multimodal cultural graphs—advancing the vision of a European Music Observatory as a complement to the European Cultural Heritage Cloud.

Project History

Download our feasibility study from https://doi.org/10.5281/zenodo.6427514.

In 2020, we studied the effect of increasingly AI-driven streaming platforms on the Slovak music repertoire. We realised that music is being recommended by algorithms, not only to YouTube or Spotify subscribers but also to radio editors, concert promoters, and festival organisers. We realised that rights management organisations (CMOs), music information centres (MICs), and music libraries, archives, and documentation centres (MLs) must change their practices to remain competitive and visible.

In the Open Music Europe project¹, we are building a data infrastructure that aims to coordinate music knowledge (in our demonstration example, Slovak music) stored in various institutional silos and systems. We wanted to “plug in” the database of the MIC Slovak Music Centre (MCS) into a global data system like Wikipedia or Spotify and, at the same time, connect it with the Slovak CMO SOZA, the Slovak National Library and the music section of the ML Bratislava City Library. A data science company, Reprex, oversees this process.

We want to ensure that a MICs and CMOs have the most accurate information about music. Whenever data is missing or new information has yet to reach their database, we try to look up the data from reliable sources and ensure that it arrives automatically in their system (to be reviewed by a knowledgeable human curator).
Make all music in Slovakia visible on all global music systems, encyclopaedia, Europeana and the European Cultural Heritage Cloud. Enable people to locate sheets of works for sale or public lending and show where people can listen to the music in various formats.
Provide the information in a dual format: enable a MIC to provide machine-readable, standardised, RDF annotated data to directly support the recommender systems of Spotify, YouTube, Deezer and other platforms.
Provide CMOs with a framework to make their members more visible to digital streaming platforms by enrichment, control and processing of available metadata and to also inform radio stations about the potential recordings that count into their local content, thus promoting local CMO members.

As seen from the points made above, this data infrastructure is beneficial to a great degree for all the parties involved.

European Interoperability Framework for Music

Nowadays most data is downloaded and used by intelligent software agents on various digital services and platforms. To meet the modern service requirements of 2025, we needed to implement an enhanced implementation of the European Interoperability Framework (EIF) that extends to privately held data. We need to communicate information and metadata about Slovak music in a way that musicologists, musicians and their managers, copyright law practitioners, citizen scientists as well as software agents understand worldwide.

The EIF is a four-layered specification for how to create world-class digital public services. In music, however, the private sector is vital, too. CISAC’s members, the CMOs that register new works, are private parties. To add weight, digital streaming platforms ingest more metadata about music in a week than Europeana Sound over ten years.

On the European Interoperability Framework see: (European Commission 2017).

To ensure legal and organisational interoperability, the members of the Open Music Consortium (SOZA–Slovak Performing and Mechanical Rights Society, MCS, Reprex) signed a Memorandum of Understanding with the Slovak Ministry of Culture; later, MCS, SOZA and Reprex signed a more technical MoU with the Slovak National Library².

The aim at the legal level was to understand the different rules of business and statistical confidentiality, as well as data protection rules in general. Harmonising the GDPR across a public information body like MCS and a private entity like SOZA is particularly challenging.
On the organisational level, we must ensure that different organisations (MCS, SOZA and the national library) understand each other’s data and workflows, for example, in identifying a musical work or a musical group with no legal personality.
On the semantic level, we must ensure that the databases of MCS, SOZA, and the national library understand each other.
The technical level must ensure that harmonised workflows result in successful data exchanges, given the legal constraints and semantic definitions.

In our system design, we followed two important new European initiatives. In connecting cultural data, European Collaborative Cloud for Cultural Heritage (ECCH), and the Feasibility study for the establishment of a European Music Observatory. Concerning connecting authoritative data across governmental and private systems, we adhere to the novel European data governance regulations³.

A data (sharing) space is a system that integrates data whenever it is needed, or is permitted, with some labour-intensive aspects of data integration postponed until it is possible to carry them out. In a data sharing space, like Reprex’s Motion Picture Dataspace, we know where the data is in each organisation, we know its format, documentation standards, even known problems and issues. But we only load them into an application, for example, an application that calculates environmental effects or GHG emissions when it is needed and permitted. Our dataspace reduces the work in such scenarios by by setting up automatic matching and mapping generation techniques⁴.

Wikibase is the software that runs the worlds largest open knowledge graph database, Wikidata, which synchronises knowledge with about 330 (different language) Wikipedias and countless national libraries, music services, and other data sources. Our dataspace is powered by an extended and configured Wikibase system. Wikibase has often been used for authority control; it is also the system of the EU Knowledge Graph⁵.

The most notorious problem plaguing digital services and databases is the unresolved issue of named entity disambiguation. With digital services providing access to over 100 million recordings, we often find dozens of artists or works with the same title. Named-entity ambiguity is confusing for humans and for autonomous AI systems, too.

Our system focuses on harmonising the registration processes of various knowledge organisations, such as libraries and CMOs. We want to ensure that every artist, music group and the manifestation of their work in music recordings and sheets receive a globally unique and persistent identifier. We want to provide new recordings and sheets in at least one public library with a globally used VIAF (and, when applicable, an ISNI) identifier. To support named-entity disambiguation going back to hundreds of years of music creations and events diaries, we are developing trustworthy AI systems that remain strictly under human control in line with the new requirements of the EU’s AI Actt⁶.

Disseminating information

Returning to our Feasibility study, we realised that about half of the music streams had data problems. They resulted in late or missed royalty payments. We also realised that almost a fifth of the Slovak repertoire had such low-quality metadata that music recommender systems could not make sense of what their recordings work. We are changing this situation. We want to ensure that we realise if the “Internet knows” misleading information about a work or recording. Using modern data-enriching techniques, we ensure that the MCS and SOZA have the most accurate 360° views of the repertoire they represent. By data dissemination, we also guarantee that digital streaming services and search engines are informed about precise and up-to-date knowledge.

We carry out health checks on the metadata of artists and labels and highlight missing, potentially misleading, or erroneous information.
Utilising Wikidata and Wikipedia, we disseminate the correct information to streaming services and search engines (which rely on knowledge graph technologies and regularly crawl Wikidata.) We are hosting a Wikipedian in Residence to find better cooperation with the Wikipedian community and the curator community of Wikidata.
We are building the Slovak Comprehensive Music Database (SKCMDb) that will inform radio stations about the potential recordings that count into their local content (legally stipulated) broadcasting quota, where orchestras can find for purchase or public lending scores of works or where professional and enthusiast audiences find any music ever made in Slovakia.
We are designing a new digital distribution model for non-profits and self-releasing artists who need help receiving professional service due to the low commercial value of their culturally valuable works.
We are also experimenting with data bias checks to find out about potential algorithmic biases that work against the Slovak repertoire.

Our Demo

Slovak Comprehensive Music Database (working demo)

Context: Open Music Observatory

You can read the long-form documentation of our plans here.

Our ambition with the development of the Open Music Observatory is to provide the technological basis and a practical roadmap for creating a European Music Observatory in a bottom-up, decentralised way. Instead of waiting for a grand, central agreement on what should a European music observatory be collecting and who should control it, we suggest a pragmatic approach: allow any data owners and collectors who satisfy certain quality and cooperation rules to add their data to an Open Music Observatory; when it reaches a sufficient maturity for use in Europe, then decide if its maintenance requires a new institutional form or not. You can read or regularly updated progress report on this work here.

Creating the Open Music Observatory is a cornerstone task of the Open Music Europe (OpenMusE) – An Open, Scalable, Data-to-Policy Pipeline for European Music Ecosystems (Open Music Europe 2023) Horizon Europe research and innovation project. This task is running till the end of the project (31 December 2025) with the collection, processing, and dissemination of more data and providing innovative, new data services in line with our exploitation pathways. This report is an accompanying document for the creation of Open Music Observatory as a digital infrastructure on the World Wide Web.

The Open Music Observatory is a digital service provider for the music industry that follows the European Interoperability Framework (EIF) definition for such services with a unique governance model. The governance model and the digital service infrastructure represent a unique innovation that considers many good examples from the European Union and other industries.

An observatory has traditionally been a permanent location for observing terrestrial, marine, or celestial events. In the past 30 years, it has also been used for long-term digital data collection programs for markets, social sciences, and humanities. Our milestone requires the start of this observatory after a lengthy and intensive planning and prototyping phase. It can be seen as a modern reimagination of the data observatory model, or the observatory 2.0. We created a new observatory model that fully aligns with the European Interoperability Framework but extends the governance of the digital services beyond public bodies, and allows the creation of a public-private partnership to manage the observatory.

We were informed and influenced by the creation of Europeana (which started out from a similar collaborative project) and their new plans to extend their digital services into the European Collaborative Cloud for Cultural Heritage (ECCCH). We aim fully interoperability with Europeana and ECCCH, but we also bring a new element into their thinking. While they are mainly aggregating the work of public sector memory institutions, we are building a governance model that allows a more successful cooperation among the private sector and the public music sector.

By the end of 2025, we aim to create an “observatory 3.0”, which already hosts many intelligent data improvement technologies and fuels innovative applications/services in line with our project’s exploitation pathways. These services are at different maturity levels, but they could not be brought to a testable MVP without building out the minimal digital infrastructure and governance model at this milestone.

References

Antal, Dániel. 2020. ‘Feasibility Study on Promoting Slovak Music in Slovakia & Abroad’. https://doi.org/10.5281/zenodo.6427514.

Antal, Dániel, Michal Grochal, and Christos Varvantakis. 2024. ‘Building a Music Data Sharing Space with Wikibase’. Zenodo. https://doi.org/10.5281/zenodo.8046977.

Antal, Dániel, Ádám Lázár, and Andor Kornél Barát. 2024. ‘Open Music Observatory. Progress Report’. Digital Music Observatory. https://doi.org/10.5281/zenodo.11564114.

Bianchini, Carlo, Stefano Bargioni, and Camillo Carlo Pellizzari di San Girolamo. 2021. ‘Beyond VIAF Wikidata as a Complementary Tool for Authority Control in Libraries’. Information Technology and Libraries 40 (2). https://doi.org/10.6017/ital.v40i2.12959.

BVDA. 2023. ‘Data Sharing Spaces and Interoperability’. Edited by Antonio Kung (Trialog), Ray Walshe (DCU), and Rigo Wenning (ERCIM). BVDA.

Curry, Edward. 2020. ‘Dataspaces: Fundamentals, Principles, and Techniques’. In Real-Time Linked Dataspaces: Enabling Data Ecosystems for Intelligent Systems, 45–62. Cham: Springer International Publishing. https://doi.org/10.1007/978-3-030-29665-0_3.

D’Ignazio, Catherine, and Lauren F. Klein. 2020. Data Feminism. Strong Ideas. Cambridge, MA, USA: MIT Press. https://data-feminism.mitpress.mit.edu/.

Diefenbach, Dennis, Max de Wilde, and Samantha Alipio. 2021. ‘[Wikibase as an Infrastructure for Knowledge Graphs: the EU Knowledge Graph]{.nocase}’. In ISWC 2021, 12922:631–47. Online, France: Springer. https://doi.org/10.1007/978-3-030-88361-4\37.

Directorate-General for Research and Innovation (European Commission), Pere Brunet, Livio De Luca, Eero Hyvönen, Adeline Joffres, Peter Plassmeyer, Martijn Pronk, Roberto Scopigno, and Gábor Sonkoly. 2022. [Report on a European collaborative cloud for cultural heritage: ex – ante impact assessment]{.nocase}. LU: Publications Office of the European Union. https://data.europa.eu/doi/10.2777/64014.

ESS. 2017. ‘Position Paper on Access to Privately Held Data Which Are of Public Interest’. European Statistical System. https://ec.europa.eu/eurostat/documents/13019146/13346094/ESS+Position+Paper+on+Access+to+privately+held+data+final+-+Nov+2017.pdf/6ef6398f-6580-4731-86ab-9d9d015d15ae?t=1511447619000.

———. 2022. Privately Held Data Communication Toolkit. 2022nd ed. Manuals and guidelines. Luxembourg: Publications Office of the European Union.

EU 2019/1024 Open Data Directive. 2019. Directive (EU) 2019/1024 of the European Parliament and of the Council of 20 June 2019 on Open Data and the Re-Use of Public Sector Information. Official Journal of the European Union. Vol. OJ L. http://data.europa.eu/eli/dir/2019/1024/oj.

EU 2022/868 Data Governance Act. 2022. ‘[Regulation (EU) 2022/868 of the European Parliament and of the Council of 30 May 2022 on European data governance and amending Regulation (EU) 2018/1724 (Data Governance Act) (Text with EEA relevance)]{.nocase}’. EUR-Lex. http://data.europa.eu/eli/reg/2022/868/oj.

EU 2024/1689 Artificial Intelligence Act. 2024. ‘[Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)]{.nocase}’. OJ OJ L 2024/1689 (July): 1–144. http://data.europa.eu/eli/reg/2024/1689/oj.

European Commission. 2017. ‘European Interoperability Framework – Implementation Strategy. Communication from the Commission to the European Parliament the Council the European Economic and Social Committee and the Committee of the Regions’. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52017DC0134.

European Commission, and Directorate-General for Communications Networks, Content and Technology. 2019. Ethics Guidelines for Trustworthy AI. Publications Office of the European Union. https://doi.org/10.2759/346720.

European Commission, Directorate-General for Education, Youth, Sport and Culture, M Clarke, P Vroonhof, J Snijders, A Le Gall, B Jacquemet, et al. 2020. Feasibility Study for the Establishment of a European Music Observatory : Final Report. Publications Office. https://doi.org/doi/10.2766/9691.

Fagerving, Alicia. 2023. ‘Wikidata for Authority Control: Sharing Museum Knowledge with the World’. Digital Humanities in the Nordic and Baltic Countries Publications 5 (1): 222–39. https://doi.org/10.5617/dhnbpub.10665.

Hudobné centrum, Reprex B.V., and Slovenská národná knižnica, and Slovenský ochranný zväz autorský. 2024. ‘Memorandum o porozumení vo vzťahu k Slovenskej súhrnnej hudobnej databáze]’.

Ministerstvo kultúry SR, and Open Music Europe. 2023. ‘Memorandum o porozumení o využití výsledkov analýz otvorených politík v kontexte slovenského kultúrneho a kreatívneho priemyslu a sektorových verejných politík v spolupráci s konzorciom pre výskum a inovácie s názvom OpenMuse. $$Memorandum of Understanding on utilizing the Open Policy Analysis results of the OpenMuse Research and Innovation Consortium in the context of Slovak cultural and creative industries and sectors’ public policies$$’. https://www.crz.gov.sk/zmluva/7645338/.

Open Music Europe. 2023. ‘[Open Music Europe (OpenMusE) – An Open, Scalable, Data-to-Policy Pipeline for European Music Ecosystems]{.nocase}’. https://doi.org/10.3030/101095295.

This project has received funding from the European Union’s Horizon Europe, research and innovation programme, under Grant Agreement No. 101095295 (Open Music Europe (OpenMusE) – An Open, Scalable, Data-to-Policy Pipeline for European Music Ecosystems (Open Music Europe 2023)). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them. ↩︎
See the references (Ministerstvo kultúry SR and Open Music Europe 2023; Hudobné centrum et al. 2024) ↩︎
See on the ECCH (Directorate-General for Research and Innovation (European Commission) et al. 2022) and the European Music Observatory (European Commission et al. 2020); on connecting the (public) European Statistical System with privately-held data: (ESS 2017, 2022); and as a general legal framework the Open Data Directive and the Data Governance Act (EU 2019/1024 Open Data Directive 2019; EU 2022/868 Data Governance Act 2022). ↩︎
You can read more in less technical language about this approach in Dataspaces: Fundamentals, Principles, and Techniques (Curry 2020), and in a more technical specification in Data Sharing Spaces And Interoperability (BVDA 2023). Reprex is an affiliated member of the Big Data Value Association (BVDA). ↩︎
The knowledge graph of the European Union: (Diefenbach, Wilde, and Alipio 2021); use for cultural authority control: (Bianchini, Bargioni, and Pellizzari di San Girolamo 2021; Fagerving 2023). Our solution: (Antal, Grochal, and Varvantakis 2024). ↩︎
Ethics guidelines for trustworthy AI (European Commission and Directorate-General for Communications Networks, Content and Technology 2019); Artificial Intelligence Act (EU 2024/1689 Artificial Intelligence Act 2024). We were greatly informed by Data feminism (D’Ignazio and Klein 2020) on the intersectional escalation of data biases that leads to disadvantages to women, small countries, or independent repertoires. ↩︎

The Finno-Ugric Data Sharing Space

Tue, 19 Nov 2024 17:16:00 +0100

For a better readability, you can acces this document in HTML, PDF, and EPUB here.

We started experimenting with the legal, organisational, semantic and technical challenges of creating a genuinely trustworthy, AI-supported data-sharing space that can find and connect tangible and intangible elements of the Finno-Ugric cultural universes. We were also seeking a better governance model for oversight for the custodians of these endangered, shrinking universes in their language and with little technical knowledge, partly as alternatives to the established Wikipedia to the open knowledge incubation method for small linguistic minorities.

Motivation and Background
Prototype for PPP Knowledge Incubation Into the Wikimedia Ecosystem
- Competency Questions: All Layers of the European Interoperability Framework Applied
- Competency Questions: Involving Wikipedians and the Wikimedia Ecosystem
Policy Alignment and Ethical Governance
References

Motivation and Background

Please check out the demo version of the Finno-Ugric Dataspace or read the long-form project description.

The Finno-Ugric Data Sharing Space builds on a previous initiative that unified all known music ever recorded or produced in Slovakia (Ministerstvo kultúry SR and Open Music Europe 2023), in collaboration with a Wikipedian-in-residence. Although Slovak music represents a relatively compact cultural domain, the relevant data was dispersed across numerous public and private institutions—each with its own metadata conventions, languages, and access protocols. This earlier effort showed that even within a narrowly defined area, knowledge integration is complex and labour-intensive—but can deliver substantial impact when institutional silos are bridged and practices aligned.

We now extend this approach into a far more fragile and urgent context: the cultural and linguistic ecosystems of endangered Finno-Ugric communities. Here, we are not simply harmonising open datasets—we are engaging with disappearing cultural universes. These are communities, such as the Livonians or Mari, where the number of fluent speakers and active custodians may be in the dozens. In many cases, the knowledge at risk has never been systematically documented, or exists only in languages and formats that are inaccessible to the communities themselves.

This dual challenge—of endangered knowledge and endangered language—calls for more than technical aggregation. Our platform must support both language-independent conceptualisation (structured metadata, RDF, linked open data) and language-bound revitalisation (cultural narratives expressed in community languages such as Livonian or Seto). A single cultural object—a song, a garment, a photograph—must exist both as semantically linked data and as a living narrative, grounded in linguistic and cultural specificity.

Although our focus is on Finno-Ugric communities, the underlying model is widely applicable. Many domains—such as interwar Slovak modernism or regional Sámi ethnographies—can be described as high-friction knowledge spaces: editorially underdeveloped, institutionally fragmented, and costly to document.

The Finno-Ugric Data Sharing Space provides a blueprint for addressing these challenges. It is a staging environment, interoperable with Wikimedia platforms, where underrepresented communities can collaboratively curate, annotate, and translate their cultural knowledge—on their own terms—prior to integration into the global commons.

Our project idea began to take shape during the Wikimedia CEE 2024 Meeting, where we discussed with Finnish, Estonian, and Hungarian Wikipedians how to strengthen the participation of Finno-Ugric minority communities in Wikimedia projects.

Our project idea began to take shape during the Wikimedia CEE 2024 Meeting, where we explored with Finnish, Estonian, and Hungarian Wikipedians how to better support participation from minority communities within the Finno-Ugric language group in Wikimedia projects after presenting our data sharing space solution powered by the Wikibase Suite.

While Wikipedia plays a central role in global knowledge-sharing, it is governed by general-purpose policies—such as notability criteria—that can unintentionally marginalise the perspectives, individuals, and artefacts of smaller communities. The narrower and more culturally specific the universe, the more likely it is that meaningful people, events, or objects will fall below notability thresholds, despite their deep significance in local contexts.

To address this, our platform introduces a curatorial layer, currently named WikiGLAM, that enables the creation of virtual exhibitions and contextual collection pages. These pages are not designed to replace Wikipedia articles, but to complement them—providing a format where cultural specificity and community storytelling can thrive, even when individual entries do not meet notability requirements.

For example, while Wikipedia may not support a standalone article for every Seto singer, a digital exhibition titled “The Seto Singing Tradition” could present ten singers, each with images, audio recordings, and biographical context—linked through Wikidata and hosted via Wikimedia Commons. This curatorial model rebalances representation for cultural minorities without compromising Wikipedia’s editorial standards.

Our unfunded project was presented both as a conference paper and a poster at the Digital Dreams and Practices conference (Digital Humanities in the Nordic & Baltic Countries, 2025). A peer-reviewed publication based on this work will be available later this year. You can download the poster in pdf format.

Prototype for PPP Knowledge Incubation Into the Wikimedia Ecosystem

Competency Questions: All Layers of the European Interoperability Framework Applied

To guide the design and implementation of the Finno-Ugric Data Sharing Space, we define the following key competencies the platform must address:

Q1: Define the Scope of the Cultural Universe

The platform is not intended to replicate a general-purpose encyclopedia. For communities like the Liv, with only a few hundred remaining speakers, the informational needs center on the preservation of their specific cultural world: geographic features bearing Liv names, remaining Liv-speaking villages and buildings, garments, oral traditions, folk songs, and the vocabulary needed to describe these in the Liv language. Our system must support the creation and structuring of a domain-specific encyclopedic resource, rooted in the cultural priorities of the communities it serves.

The Finno-Ugric Dataspace should support domain-specific encyclopedic content generation that reflects the cultural worldview and linguistic nuance of the Livonian community, rather than defaulting to general-purpose geographic knowledge as seen in standard Wikipedias. This content becomes the foundation of a Livonian-language encyclopedic resource—complementary to but distinct from global projects.

Q1: generate structured entries

Can the system generate structured entries and multilingual encyclopedic descriptions of culturally significant Livonian geographical locations, using the Livonian Place Name Catalogue as a source, and provide historical variants in Latvian, Russian, and German where available?

From the Livonian Place Name Catalogue, the place Irē (Livonian) / Mazirbe (Latvian) is one of the most significant locations in Livonian cultural memory. The system should be able to: - Represent the Livonian name (Irē), along with Latvian (Mazirbe), and historical Russian and German variants (Мазирбе, Maserwieken). - Link the place to its role in Livonian history and cultural revival (e.g., site of the Livonian House of the People). - Include geographical coordinates and administrative details. - Enable users to generate a short encyclopedia-style page or entry in Livonian, describing Irē’s cultural and linguistic relevance. - Link to associated artefacts, songs, photos, or oral histories from this location.

Q2: Bridge Institutional and Linguistic Silos

Knowledge about stateless or minoritized communities—such as the Seto, Mari, or Liv—is distributed across public and private collections in countries with different national languages (Latvian, Estonian, Finnish, Hungarian, Russian, German). These collections are managed under varied knowledge organization models typical to GLAM institutions. Our platform must support cross-lingual and cross-institutional discovery, enabling users to locate and contextualize Seto or Liv heritage materials scattered across multiple metadata regimes and languages.

The goal of the Finno-Ugric Data Sharing Space is to act as a cross-lingual, cross-domain discovery layer, allowing communities and researchers to access cultural heritage materials scattered across institutions and described in different vocabularies and languages. This enables community members, even without formal archival or linguistic training, to meaningfully engage with their heritage.

Q2 Can the system enable a user—regardless of their language proficiency—to discover and access bibliographic records, archival descriptions, and museum artefacts?

Can the system enable a user—regardless of their language proficiency—to discover and access bibliographic records, archival descriptions, and museum artefacts related to traditional Seto clothing from the 19th and 20th centuries, even when those records are stored in different institutions and described in Estonian, Russian, or German metadata schemas?

A user is researching Seto traditional women’s clothing, including:

The silver brooches and layered aprons worn during ceremonial events;
Garments documented in photographs or field sketches from 19th-century ethnographic expeditions;
Mentions in ethnographic bibliographies, published in Estonian or German;
Museum entries* in databases like MuIS (Estonian Museums Information System) where items are tagged with terms like “Setu rahvarõivad”, “Setu naise kostüüm”, or “женская одежда сето”.

The system should:

Recognize synonyms and multilingual terms (e.g., Setu = Seto, naise rahvarõivad = women’s folk clothing);
Normalize metadata differences across domains (e.g., MARC for bibliographies, ISAD(G) or RiC for archives, CIDOC CRM for museum artefacts);
Return a harmonized set of entries with linked metadata;
Optionally generate a multilingual summary of the findings and enable further semantic enrichment.

Q3: Translate Information into Structured, Interoperable Knowledge

Our system must be able to convert both structured (e.g., datasets, catalogue entries) and unstructured (e.g., books, photos, oral histories) information into semantic, RDF-based knowledge representations. To ensure cross-domain compatibility, we adopt a composite data model drawing from established ontologies: DCTERMS (libraries), Records in Contexts (archives), CIDOC CRM (museums), and copyright ontologies (for streaming platforms and contemporary media). For example, a museum photograph of Seto traditional clothing must be described in a language-agnostic, conceptually robust format that links garments, places, people, events, and creators.

Q3 Etract structured RDF-based statements from mixed-format cultural heritage records

Can the system extract structured RDF-based statements from mixed-format cultural heritage records—such as a museum photograph with a descriptive title and date—and represent them using interoperable, language-independent vocabularies that support integration across library, archive, and museum domains?

Example objects: MuIS Record: Setu mehed Võmmorski külas Setomaa vallas (1913)
Title (Estonian): Setu mehed Võmmorski külas Setomaa vallas
Date: 1913
Institution: Eesti Rahva Muuseum (Estonian National Museum)
Item type: Photograph
Language of metadata: Estonian
Collection domain: Museum (but relevant to archives and cultural ethnography)

From this item, the system should derive:

There exists a photograph as a physical/visual object.
The photograph was created in 1913.
The photograph depicts Seto men in Võmmorski village, in the Setomaa region.
Since the photo is of people in 1913 and is not marked “nude,” it implicitly depicts traditional Seto male clothing.
The image is part of a museum collection.
The location and people depicted have cultural and ethnographic significance.

@prefix crm: <http://www.cidoc-crm.org/cidoc-crm/> .
@prefix dcterms: <http://purl.org/dc/terms/> .
@prefix schema: <http://schema.org/> .
@prefix ex: <http://example.org/entities/> .

ex:photo_610034 a crm:E22_Man-Made_Object ; # The photograph as an object
 crm:P2_has_type ex:Photograph ;
 crm:P4_has_time-span ex:year_1913 ;
 crm:P108i_was_produced_by ex:photo_creation_event_1913 ;
 crm:P62_depicts ex:seto_men_group_1913 ;
 crm:P7_took_place_at ex:voemmorski_village ;
 dcterms:subject ex:seto_men , ex:seto_traditional_male_clothing ;
 dcterms:creator ex:estonian_national_museum ;
 schema:inLanguage "et" ;
 dcterms:isPartOf ex:muis_collection .

ex:photo_creation_event_1913 a crm:E12_Production ;
 crm:P4_has_time-span ex:year_1913 ;
 crm:P7_took_place_at ex:voemmorski_village .

ex:seto_men_group_1913 a crm:E21_Person ; # or an instance of E74_Group if individuals not known
 crm:P131_is_identified_by "Setu mehed"@et ;
 crm:P74_has_current_or_former_residence ex:voemmorski_village .

ex:seto_traditional_male_clothing a crm:E28_Conceptual_Object ;
 crm:P2_has_type ex:Traditional_Clothing ;
 crm:P70i_is_documented_in ex:photo_610034 .

ex:voemmorski_village a crm:E53_Place ;
 crm:P87_is_identified_by "Võmmorski küla"@et ;
 crm:P89_falls_within ex:setomaa_region .

Q4: Translate Structured Knowledge into Human-Readable Language

Many small communities lack trained librarians, information professionals, or technical ontologists. Therefore, we must offer tools and workflows that can convert structured data into natural-language descriptions in Liv, Mari, Seto, or other relevant languages. Using computational linguistics and native speaker input, we aim to generate readable, accurate sentences—enabling the production of culturally grounded encyclopedic content, and empowering community members to reclaim access to knowledge historically fragmented and dispersed across external archives.

Expected features

✅ Data grounding: Every part of the sentence can be mapped back to a specific RDF triple or identifier.
✅ No hallucination: The output does not introduce facts not present in the data model.
✅ Domain-appropriate naturalization: The sentence respects cultural terms (“Seto men”, “Võmmorski village”) and avoids overly technical phrasing.
✅ Reference transparency: The institutional source (ERM), identifier, and URI are clearly included.

Q4 Convert structured RDF-based knowledge to culturally appropriate, and verifiable sentences

Can the system convert structured RDF-based knowledge—linking people, places, dates, and artefacts—into clear, culturally appropriate, and verifiable sentences in natural language, such as Livonian, Seto, Mari, or English, grounded in identifiable knowledge items and their provenance?

To offer community members, editors, or citizen curators a tool that:

Selects a set of RDF nodes (e.g., one photo, one person, one location)
Uses predefined or AI-augmented linguistic templates for generating narrative
Outputs fully traceable text in multiple target languages
Allows human validation or post-editing by native speakers

Example Workflow (based on Q3 TTL output):

Structured Input (summarized): - There is a photograph (ex:photo_610034) - It depicts a group of Seto men (ex:seto_men_group_1913) - It was taken in 1913 (ex:photo_creation_event_1913) - In Võmmorski village (ex:voemmorski_village) - It’s part of the Estonian National Museum collection - With item ID: ERM Fk 213:172 - Permanent link: https://opendata.muis.ee/object/610034

Target Natural Language Output (English):

“This photograph, taken in 1913 in Võmmorski village (Setomaa), shows a group of Seto men and is held in the Estonian National Museum under the catalogue ID ERM Fk 213:172. The record is available at https://opendata.muis.ee/object/610034.”

Q5: Enable Legal and Technical Interoperability Across Public and Private Knowledge Sources

Much of the cultural heritage of small and stateless communities is held outside formal institutions—in private collections, family archives, grassroots NGOs, and online platforms. While significant effort has been invested in connecting public libraries, archives, and museums, our platform must also support legal and organizational frameworks that enable data exchange and enrichment between public and private stakeholders. This includes respecting intellectual property rights, cultural sensitivities, and varying levels of openness, while still enabling meaningful integration with Wikidata, Wikimedia Commons, and other open repositories. Our goal is to create an inclusive data-sharing model that recognizes the distributed nature of cultural custody in small communities.

To goal of or Finno-Ugric Data Sharing Space offer legally respectful, semantically rich representation of cultural materials sourced from non-institutional domains—especially those that are digital-first, grassroots-curated, and culturally vital, but fall outside traditional GLAM pipelines. It must make this data linkable and referenceable, even if full reuse is not permitted—thus reflecting the true complexity of the heritage landscape.

Can the system document, enrich, and share information about community-curated cultural materials—such as playlists of contemporary Mari, Udmurt, or Samoyedic music—while respecting intellectual property rights, platform-specific terms of use (e.g., Spotify API rules), and the data-sharing preferences of private contributors or NGOs, and still enable integration with open knowledge platforms like Wikidata or Wikimedia Commons where legally and ethically possible?

Example: The Hõimulõimed Music Collection

A Set of playlists curated by the Hõimulõimed association includes:
Contemporary Mari folk-pop tracks available via Spotify
Experimental Udmurt and Samoyedic music hosted on Bandcamp
Unreleased or privately circulated tracks contributed by local musicians or families

System Responsibilities:

Respect platform licensing

For Spotify, adhere to Spotify’s Developer Terms when pulling metadata via the API
For Bandcamp, respect artist-defined permissions (e.g., embeds allowed vs.not)

Respect NGO & community rights

Hõimulõimed might retain curatorial or community-sourced contextual information
Metadata (e.g., language, artist background, cultural function of the song) should be attributed to them, with a Creative Commons or custom community license as applicable

Semantic alignment with archival and Wikidata models

Describe each track using elements from DCTERMS, FRBRoo, or Schema.org (e.g., dcterms:creator, schema:inLanguage, schema:genre)
Link playlists as curated thematic collections (e.g., schema:isPartOf → “Contemporary Mari Folk Sounds”)

Access control

materials unavailable for full reuse (e.g., non-commercial Spotify tracks) can be indexed, not embedded
For Wikimedia, only materials with compatible licensing (e.g., CC BY-SA) are directly transferred to Commons or Wikidata

Competency Questions: Involving Wikipedians and the Wikimedia Ecosystem

Q6: Integrate Contemporary and Non-Traditional Cultural Expressions

In many cases, the most vibrant and visible expressions of cultural identity are not found in traditional heritage institutions, but on commercial platforms and personal networks. For example, playlists of Mari folk-punk, Udmurt folktronica, or Seto death metal circulate primarily through Spotify and YouTube—not museums or archives. Our system must be able to recognize, document, and link to these contemporary cultural artifacts, treating them as part of the evolving knowledge base. This means supporting metadata ingestion from platforms like Spotify, retaining links to their rich descriptive ecosystems (e.g., genre, lyrics, release context), and providing tools for community annotation, translation, and contextualization.


	The following collections contain musical works that are almost always sung in the given language: Khanty Mansi Musical Works Collection; Samoyedic Musical Works Collection; Livonian Musical Works Collection; Veps Musical Works Collection ;Saami Musical Works Collection; Komi Musical Works Collection; Hungarian Musical Works Collection; Finnish Musical Works Collection; Mari Musical Works Collection; Udmurt Musical Works Collection; Estonian Musical Works Collection; Erzya Moksha Musical Works Collection

Playlists

To goal of the FUD is to empower Wikipedians, citizen scientists and activists to contribute culturally and linguistically meaningful content that is is ethically shared and distributed, and which is curated collaboratively into language preservation resources, including dictionaries, lexeme repositories, and encyclopedic entries.

We also want to contribute to provide best use examples of the Wikimedia Commons, Wikidata, the Wikibase Suit, but even more the fledgling Lexeme extension and Wiktionary sister project, because the use of the Lexeme extension is currently lacking really good practices of integration with mainstream computational linguistic projects, and remains a bit cumbersome to work with even with people familar to the Wikibase Cloud or Suit solutions.

Q6 Maningfully transform community-generated cultural expressions into structured knowledge

Can the system meaningfully integrate contemporary, community-generated cultural expressions—such as photographs, videos, oral explanations, and audio recordings—contributed by Liv, Seto, or Mari activists and citizen scientists, and transform these materials into structured, multilingual knowledge that supports Wikimedia’s broader mission in language and cultural preservation?

A Seto speaker uploads to Wikimedia Commons:

A recording of a traditional song, including an oral explanation of its meaning
A short video explaining how a ritual skirt is folded and worn
A written caption or story in Seto, containing several words not yet documented in Wiktionary or Lexeme

The system should:

Ingest and semantically link the media (e.g., via schema:subject, schema:inLanguage)
analyse the spoken and written text to identify potential new lexemes
With expert review, generate: Wikidata Lexemes with grammatical properties (e.g., verb forms, declensions); Wiktionary entries with definitions, pronunciation, example usage.
Link each lexeme back to its attestation context (the song, the ritual, the speaker), preserving linguistic richness and cultural grounding
Allow these outputs to feed back into future encyclopedic articles or WikiGLAM exhibitions in Liv, Seto, Mari, or multilingual format.

Q7: Use Computational Linguistics to Enrich Lexicographic Resources via Lexeme and Wiktionary Integration

The Liv, Seto, and Mari languages are primarily oral, with limited formal grammars or lexicons. Much of the linguistic content connected to cultural heritage—such as song lyrics, object descriptions, or oral histories—is undocumented or inconsistently recorded. Our platform aims to address this gap by using computational linguistic tools to extract structured linguistic data from community-provided texts and audio.

Rather than asking citizen linguists to populate Wikidata Lexemes or Wiktionary pages manually, we design a workflow that enables automated or semi-automated population of these resources. Community narratives, captions, or oral recordings connected to cultural artefacts (e.g., traditional garments, buildings, rituals) are analysed for vocabulary, morphology, and syntax, and the outputs are formatted for compatibility with the Lexeme extension of Wikibase as well as Wiktionary standards.

This approach does several things:

Improves linguistic documentation of small and endangered languages in structured, machine-readable formats;
Links lexemes directly to cultural context, grounding abstract vocabulary in lived practice;
Bridges community knowledge with Wikimedia’s structured linguistic ecosystem, making small languages more visible and semantically integrated.
By doing so, we make the Lexeme extension not just a passive repository, but an active interface between cultural knowledge, language preservation, and structured knowledge production—lowering the barrier of participation while improving linguistic data quality.

To turn cultural expressions from Q6 into linguistically structured outputs that:

Populate and enrich Wikidata Lexeme and Wiktionary
Strengthen formal grammatical understanding of oral or underdescribed languages
Provide community-owned language resources that support both preservation and active use

Q7 Extract structured lexical and grammatical data from contributions

Can the system extract structured lexical and grammatical data from culturally grounded community contributions—such as songs, oral histories, and captions—and use computational linguistics tools to semi-automatically populate Wikidata’s Lexeme extension and Wiktionary entries, supporting formal documentation and visibility of Liv, Seto, and Mari languages?

Q8: Interpretive Layers That Complement Wikipedia’s Notability Criteria

Many elements of small or specialized cultural heritages—such as regional festivals, family-held artefacts, or influential community members—do not meet the general notability criteria required for inclusion in Wikipedia, especially in larger language versions like English. However, these elements are essential to preserving and understanding the cultural world of communities like the Seto or Liv.

Our platform addresses this by supporting curated digital exhibitions and thematic collection pages, such as those developed in our proposed WikiGLAM environment. These exhibitions are not encyclopedic entries per se, but interpretive presentations that provide context, narrative framing, and cultural linkage. They allow, for example, a single page on “Seto Singing Traditions” to present multiple individuals, songs, costumes, and local histories, with appropriate links to Wikidata and Wikimedia Commons, even if no individual item is independently notable.

This competency ensures that structural inclusion gaps in Wikipedia can be ethically and meaningfully addressed without undermining community norms—by offering a parallel format for expression and representation within the broader Wikimedia ecosystem.

The Finno-Ugric Dataspace should allow and support the creation of a WikiGLAM-style virtual exhibition page, as developed by the Estonian Wikipedia, with embedding of Commons-hosted images and videos; linking each item to its Wikidata entity (e.g., a costume, a festival, a village); bilingual or multilingual narrative (e.g., Seto, Estonian, and English); Attribution to community curators, knowledge holders, and institutions, and a clear URI for citation and long-term access.

### Q8 Can the system support the creation of curated digital exhibitions?

Can the system support the creation of curated digital exhibitions—similar in form to the Müstiline ürgmeri page on Estonian WikiGLAM—that present culturally significant people, objects, and practices (such as Seto traditional costumes) through narrative, multimedia, and thematic storytelling, even when these subjects do not meet individual notability criteria for Wikipedia inclusion?

Our curator, or a museum curator wants to document the evolution of Seto women’s traditional clothing from the 19th century to the present with the inclusion of citizen scientists and Seto activists with the following content, including: - Photographs of garments from museum collections and family archives

Descriptions of how garments were worn, made, and passed down
Personal recollections or quotes from Seto women about their significance
Explanations of the symbolic meanings of specific colors or embroidery
Biographical notes on seamstresses or dancers who wore these clothes

Individually, these elements may not qualify for standalone Wikipedia pages—e.g., a single apron design or the life story of a local seamstress. But together, they form a culturally rich narrative that deserves preservation and visibility.

Policy Alignment and Ethical Governance

While the Finno-Ugric Data Sharing Space introduces technical innovations in data modeling and interoperability, its deeper aim is to restore governance of cultural knowledge to the communities it concerns. For communities like the Liv, Mari, or Seto—whose cultural heritage is widely scattered and whose languages are endangered—the ability to access, review, and represent their own cultural record is often constrained by external cataloguing practices, linguistic barriers, and institutional gatekeeping.

Our project reimagines data governance not merely as compliance with standards, but as a process of empowerment. We use policy frameworks such as the European Interoperability Framework (EIF) and the Data Governance Act (DGA) not as goals in themselves, but as tools to support a deeper ethical objective: enabling communities to reclaim agency over how their knowledge is curated, contextualized, and shared.

Making Interoperability Meaningful

The EIF is typically used to guide the design of digital public services across EU member states. We apply it in a new context: community-facing knowledge infrastructures for small-language and niche cultural domains. By aligning legal, organizational, semantic, and technical layers, we enable knowledge to flow across institutional and national borders—making it legible and actionable for the communities it describes. For example, we analyse how a traditional garment might be described differently in archival and library metadata, and we use that understanding to design interoperability that reflects both institutional logic and community meaning. (Curry 2020; EBU and Gaia-X 2022, p16; Nagel and Lycklama 2021)

Making Governance Participatory

The DGA creates legal pathways for data altruism and data intermediation, allowing privately held data to be shared for the public good under new governance models. We build on this by offering a platform where community actors can act as curators, reviewers, and contributors, even when they are not formal institutions. Our system respects provenance, attribution, and usage rights—while ensuring that data about communities is never decoupled from the communities themselves. This aligns with the DGA’s vision of lawful, ethical data sharing, but also extends it toward participatory cultural stewardship.

From Infrastructure to Oversight

Ultimately, we see interoperability and governance frameworks as enablers of ethical oversight, and frameworks of making AI-supported systems like ours trustworthy with real human control. Small and underrepresented communities must be able not only to access what is known about them, but also to critically evaluate and revise that knowledge. They must be able to question categorizations, restore missing context, add language-specific nuance, and share knowledge in the forms and formats they choose. Our staging environment makes this possible by integrating structured data, human-readable language generation, and review workflows that center community consent.

In this sense, the Finno-Ugric Data Sharing Space is not just a technical platform—it is a prototype for community-led knowledge governance in the digital age.

References

:::::::: {#refs .references .csl-bib-body .hanging-indent entry-spacing=“0”} ::: {#ref-antal_grochal_wikimedia .csl-entry} Antal, Dániel, Michal Grochal, and Christos Varvantakis. 2024. ‘Building a Music Data Sharing Space with Wikibase’. Zenodo. https://doi.org/10.5281/zenodo.8046977. :::

::: {#ref-curry_dataspaces_2020 .csl-entry} Curry, Edward. 2020. ‘Dataspaces: Fundamentals, Principles, and Techniques’. In Real-Time Linked Dataspaces: Enabling Data Ecosystems for Intelligent Systems, 45–62. Cham: Springer International Publishing. https://doi.org/10.1007/978-3-030-29665-0_3. :::

::: {#ref-dataspace_for_cci_2022 .csl-entry} EBU, and Gaia-X. 2022. ‘Dataspace for Cultural and Creative Industries. Position Paper. v.2.0’. Gaia-X. https://gaia-x.eu/wp-content/uploads/2022/10/EBU_position-paper_Media-Data-Space.pdf. :::

::: {#ref-open_music_europe_sk_mou_2023 .csl-entry} Ministerstvo kultúry SR, and Open Music Europe. 2023. ‘Memorandum o porozumení o využití výsledkov analýz otvorených politík v kontexte slovenského kultúrneho a kreatívneho priemyslu a sektorových verejných politík v spolupráci s konzorciom pre výskum a inovácie s názvom OpenMuse. $$Memorandum of Understanding on utilizing the Open Policy Analysis results of the OpenMuse Research and Innovation Consortium in the context of Slovak cultural and creative industries and sectors’ public policies$$’. https://www.crz.gov.sk/zmluva/7645338/. :::

::: {#ref-design_principles_data_spaces_2021 .csl-entry} Nagel, Lars, and Douwe Lycklama, eds. 2021. ‘Design Principles for Data Spaces. Position Paper. Version 1.0.’ Open DEI. https://doi.org/10.5281/zenodo.5244997. ::: ::::::::

Listen Local

Tue, 29 Sep 2020 10:00:00 +0000

Listen Local Slovakia

Listen Local Slovakia was an experimental programme developed during the COVID-19 period to improve the visibility of local repertoire in an increasingly algorithm-driven music ecosystem. It addressed a structural imbalance: recommendation systems trained on global-scale data tend to reinforce existing popularity, making smaller-language and local artists systematically harder to discover.

The project aimed to create alternative pathways for music discovery through locally grounded playlists, radio programming, and data-driven promotion tools. Instead of relying on opaque platform algorithms, it explored transparent, reproducible methods for connecting artists with audiences, combining open data, playlist analysis, and collaborative curation.

My role was to transfer and adapt the Hungarian “Hallgass Hazait” concept to Slovakia, redesigning it for a different linguistic, institutional, and market environment. This involved aligning stakeholders across the Slovak music ecosystem and introducing an open policy analysis approach to ensure that methods, data, and assumptions remained transparent and reusable .

The project also served as an early prototype for later work on music data infrastructures. It demonstrated that discoverability is not only a cultural or editorial issue, but a data governance problem linked to metadata quality, identifier systems, and rights information — challenges that are widely recognised across European creative industries .

For me, Listen Local Slovakia marked the transition from advocacy for local music to building the data and policy tools required to support it sustainably.

“Big data creates injustice.” – Cathy O’Neil, author of Weapons of Math Destruction

Our Listen Local project is aiming to create better radio playlisting, personal playlisting and concert promotion in a local context: within Slovakia or at a more specific level, Flanders or even the city of Utrecht or Budapest. We aim to place our partner’s music in local radio lists, personal playlists, and grow their fan base during the COVID-19 pandemic so that in 2021 they can eventually meet in the venues again, and carry out longer, more successful tours than ever.

Signing up for artists, labels, talent managers below.

What is the problem?

Most people are listening to music on Spotify or Apple Music, and at home on YouTube. They are following advice from algorithms tested on hundreds of millions of people. Anybody can create playlists, but mainly Spotify, Apple, or YouTube create the playlists with their AI algorithms and their staff and their interests.

Listen Local: Open Collaboration Experiment & Feasibility Study - how you can participate in the experiment.
Forgetify: Popular Music That Nobody Listens To - because you do not want to end up on the list of 0 streams.
Who Is Not Recommended On Spotify - you need to have 1000x more followers to double the routes leading to your artist profile and recordings with the current algorithms. We need new recommendation engines to dig out the local artists buried under a pile of international hits.
The Racist Music Algorithm — big data algorithms will increase injustice and breach social norms if they are trained on improper data, or the algorithms are biased. This is why we are working on transparent, peer-reviewed, open source alternatives.
This Is Not Democracy - Robin Kester is one of the first artists who is looking for alternative recommendation routes towards the UK and Ireland.
The Mountains Are Higher Next Door And They Understand Our Lyrics - the Twentees are flying under the radar in Slovakia, Serbia, Romania, because they sing in a minority language.
October Playlist: Wounded Egos - a short case study of a playlist analysis and comparison.

If you are an artist, you can simply fill in a form identifying your music on Spotify, filling in a few biographical data about yourself (or your band, ensemble, group). If you represent several artists as a talent manager, music publisher, record label, granting agency or music export office, please get in touch with us for onboarding.

We are aiming to create an alternative market measurement, marketing and recommendation engine tool that is based on open source software, open collaboration, full scientific transparency in an open collaboration with artists, small businesses and the entire music ecosystem.

create new playlists Let’s build up a weekly, relevant, non-exclusive playlists for music coming from Utrecht, Antwerpen, Rotterdam, Den Haag, Flanders or the Netherlands that smoothly blend in local music with relevant, spiritually or musically matching songs that fans of the Utrecht, Antwerpen, Rotterdam, Den Haag bands like.
create apps! If Forgetify can pick out songs that nobody, ever, ever listened to on Spotify, not even the performer’s mother, sister, friends, not even their enemies, than we can must create new apps that allows fans in Utrecht, visitors or Rotterdam and the virtual community of Den Haag to find relevant music, and to connect with music created in these communities.

We are a bunch of musicians, musicologists, music analysts, music journalists, producers who want to make this happen.

Data Use

From participating artists we only ask questions that they usually post on their websites or would tell to music journalists or bloggers. We will put all this information (with some clearly stated exceptions in the registration form) on the website of the experiment.
We will use similar data used to analyze this playlist about the participants’ music. This data is an analysis of their released music, not about the persons who play the music, or their earnings.
We do not ask for any financial data, and we do not have access to any financial data of the participants.

Transparency

Listen Local and the Demo Music Observatory grew out of a large, collaborative project of collective management societies, grant managers, music distributors, venues, and other music stakeholders who joined forces to collect more royalties from 2014 onwards starting with three, and eventually encompassing more than a dozen countries.

The entire budget of our project is 10,000 euros, co-funded by the Slovak Arts Council, SOZA and Consolidated Independent, a distributor of independent music. We hope that this will be enough to create a demonstration and a detailed feasibility study that makes the user needs of Slovak managers, labels, publishers and artists themselves to continue this project in 2021. The project will be carried out by Reprex, a Dutch start-up co-founded by the author of the Slovak Music Industry Report, the first Hungarian Music Industry Report, Private Copying in Croatia, and the Central European Music Industry Report 2020.
Participation for artists and music venues is free in the experiment. In 2020, we would like to find a viable business model that keeps this tool a good value for money for anybody in the independent music scenes.
We are asking labels, publishers, talent managers to contribute to our experimental budget on a crowdsourcing basis at a later stage, if they will continue to use this service in 2021. Currently our experiment is free. Signing up for artists, labels, talent managers above.

Unlabel

Tue, 29 Sep 2020 10:00:00 +0000

Unlabel is a concept and pilot developed within the Open Music Europe framework that explores how under-documented, culturally valuable music can be made visible, usable, and distributable in the digital ecosystem. It addresses a structural problem identified across European music systems: large volumes of repertoire remain “invisible” because rights, metadata, and attribution are incomplete or inconsistent .

The project builds on policy and technical work around interoperable copyright data and metadata governance, particularly the need for machine-readable, trustworthy identifiers and rights information across the lifecycle of musical works and recordings . Instead of attempting full reconstruction at source, Unlabel focuses on pragmatic repair and enrichment of legacy data, enabling works to re-enter circulation.

Developed in collaboration with the Swedish distributor Aloaded, Unlabel tests how improved metadata, identifier alignment, and lightweight rights modelling can unlock dormant catalogues for distribution platforms and AI-ready environments. It operates at the intersection of archives, collective management, and digital distribution.

For me, Unlabel represents a shift from analysis to implementation: a concrete attempt to demonstrate that better data governance is not only a policy objective, but a practical pathway to cultural reuse, market access, and long-term preservation.

Portfolios | Antal Dániel honlapja

Finno-Ugric Data Sharing Space

Udmurt Történeti Fotógyűjtemény

Spaces of a Constructive Dialogue

Spaces of a Constructive Discourse

Főfotó

WikiMuseum

Key exhibitions & outcomes

Key insights / lessons

TextileBase

Why Textiles?

Who is TextileBase for?

Examples from the Database

Connecting Artefacts Across Borders

Understanding Secondary Sources

Following Shifting Place Names

Building a Wikimuseum for Dispersed Collections

Next Steps

SKCMDb

Early Data Access

Wikibase Interfaces

SPARQL Endpoint

Sampo Semantic Browser

Next steps

Project History

European Interoperability Framework for Music

Music Data Sharing Space

Disseminating information

Our Demo

Context: Open Music Observatory

References

The Finno-Ugric Data Sharing Space

Motivation and Background

Prototype for PPP Knowledge Incubation Into the Wikimedia Ecosystem

Competency Questions: All Layers of the European Interoperability Framework Applied

Q1: Define the Scope of the Cultural Universe

Q1: generate structured entries

Q2: Bridge Institutional and Linguistic Silos

Q2 Can the system enable a user—regardless of their language proficiency—to discover and access bibliographic records, archival descriptions, and museum artefacts?

Q3: Translate Information into Structured, Interoperable Knowledge

Q3 Etract structured RDF-based statements from mixed-format cultural heritage records

Q4: Translate Structured Knowledge into Human-Readable Language

Q4 Convert structured RDF-based knowledge to culturally appropriate, and verifiable sentences

Q5: Enable Legal and Technical Interoperability Across Public and Private Knowledge Sources

Q5 Dcument, enrich, and share information about materials while respecting intellectual property rights

Competency Questions: Involving Wikipedians and the Wikimedia Ecosystem

Q6: Integrate Contemporary and Non-Traditional Cultural Expressions

Q6 Maningfully transform community-generated cultural expressions into structured knowledge

Q7: Use Computational Linguistics to Enrich Lexicographic Resources via Lexeme and Wiktionary Integration

Q7 Extract structured lexical and grammatical data from contributions

Q8: Interpretive Layers That Complement Wikipedia’s Notability Criteria

Policy Alignment and Ethical Governance

Making Interoperability Meaningful

Making Governance Participatory

From Infrastructure to Oversight

References

Listen Local

What is the problem?

Let’s Do This Together!

Data Use

Transparency

Unlabel