Heritage Data Lab

Heritage Data Lab
Name	Heritage Data Lab
Formation	2015
Type	Research group
Location	London, United Kingdom
Fields	Digital heritage, data curation, computational history
Parent organization	British Library

Heritage Data Lab is a research group focused on the digitization, curation, and computational analysis of cultural heritage collections. Combining archival practice with machine learning and geospatial analysis, the Lab works at the intersection of libraries, museums, and research universities to enable broad access to historical records and visual culture. It engages with national institutions, international consortia, and community archives to create interoperable datasets and reproducible workflows.

Overview

Heritage Data Lab operates within a network of institutions including the British Library, National Archives (United Kingdom), Victoria and Albert Museum, Museum of London, Natural History Museum, London, Tate Galleries, Science Museum, London, National Portrait Gallery, London and collaborates with universities such as University College London, University of Oxford, University of Cambridge, King's College London, University of Edinburgh, University of Manchester, University of York, University of Leicester, University of Glasgow, University of Durham, University of Sheffield, University of Birmingham, University of Liverpool, Queen Mary University of London, London School of Economics, Goldsmiths, University of London, Royal Holloway, University of London, University of Bristol, University of Exeter, University of Nottingham, University of Southampton, University of St Andrews, University of Warwick, University of Aberdeen, University of Bath, University of Kent, University of Reading, University of Sussex, University of Wales Trinity Saint David, University of Stirling, University of Hull, University of Surrey, University of East Anglia, University of Lancaster, University of Swansea, University of Ulster and research centres including the Alan Turing Institute and European Research Council-funded projects. The Lab interfaces with funders and heritage networks such as the Arts and Humanities Research Council, National Endowment for the Humanities, British Museum, ICOMOS, UNESCO, Council of Europe, Historic England, National Trust (United Kingdom), English Heritage, Heritage Lottery Fund, Wellcome Trust, Jisc, European Commission, Horizon 2020, Creative Europe, European Cultural Foundation, Getty Foundation and Paul Mellon Centre.

History

Founded in 2015 as a response to large-scale digitization efforts by institutions like the British Library, National Archives (United Kingdom), Bibliothèque nationale de France, Library of Congress, Smithsonian Institution, Biblioteca Nacional de España, Staatsbibliothek zu Berlin and Deutsche Digitale Bibliothek, the Lab traces its intellectual lineage to projects such as Europeana, Digital Public Library of America, Google Books, Project Gutenberg, British Newspaper Archive and collaborations with technology partners including Microsoft Research, Google Research, Amazon Web Services, IBM Research, Oracle and OCLC. Early leadership included scholars who previously worked at Courtauld Institute of Art, Institute of Historical Research, Renaissance Society of America, Royal Historical Society, Society of Antiquaries of London, English Heritage, Historic Scotland and the National Trust (United Kingdom). The Lab's public milestones coincided with exhibitions at the British Library, contributions to conferences like the Digital Humanities Conference, International Congress on Medieval Studies, CHI Conference on Human Factors in Computing Systems, ACM SIGIR, NeurIPS, ACL Conference, IEEE VIS Conference and policy engagements with Department for Digital, Culture, Media and Sport (UK), European Commission Directorate-General for Education, Youth, Sport and Culture and parliamentary inquiries on digital preservation.

Collections and Datasets

The Lab curates datasets drawn from collections such as the British Library's digitized newspapers, the National Archives (United Kingdom)'s records, the V&A's object metadata, the Tate Galleries' image corpus, the Science Museum, London's technical drawings, the National Maritime Museum's logbooks, the Royal Geographical Society's maps, the Historic England archives, the National Railway Museum collections, the Imperial War Museums photograph archives and local archives like the London Metropolitan Archives and Greater Manchester County Record Office. Datasets include OCRed newspapers, handwritten manuscripts transcriptions (crowdsourced with platforms like Zooniverse), geo-referenced maps, 3D models from the Victoria and Albert Museum and British Museum, oral history corpora from British Library Sounds, digitized sound collections from the BBC Archive, visual culture corpora from the National Portrait Gallery, London and annotated datasets aligned with standards such as Dublin Core, CIDOC CRM, METS, TEI, IIIF, Linked Open Data (LOD) vocabularies and Wikidata mappings.

Methodologies and Tools

Methodologies include computer vision pipelines trained on corpora from the Tate Galleries and National Portrait Gallery, London, natural language processing models fine-tuned on texts from the British Library and Library of Congress, named-entity recognition using gazetteers from the Ordnance Survey, geospatial analysis linked to Historic Environment Records, handwriting recognition models referencing trancriptions from the Bodleian Libraries and Cambridge University Library, and linked data transformation aligning with the Europeana Data Model. Tools and platforms used or developed include extensions of IIIF viewers, custom ingestion workflows for Omeka, CollectionSpace, ArchivesSpace, Drupal-based portals, APIs integrating with Wikidata and OpenRefine, machine learning frameworks such as TensorFlow, PyTorch, scikit-learn, and cloud services like Amazon Web Services, Google Cloud Platform and Microsoft Azure.

Collaborations and Partnerships

Partnerships span national institutions (e.g., British Library, National Archives (United Kingdom), British Museum), academic departments at University College London, University of Oxford, University of Cambridge, King's College London, University of Edinburgh and University of Leeds, international museums such as the Metropolitan Museum of Art, Louvre, Rijksmuseum, Prado Museum, Pergamon Museum, Hermitage Museum, National Gallery (London), Guggenheim Museum, Museo Nacional Centro de Arte Reina Sofía, Museum of Modern Art, Getty Research Institute and technology collaborators including Google Cultural Institute, Microsoft AI for Cultural Heritage, Wikimedia Foundation, Internet Archive, Zooniverse and commercial partners. The Lab has engaged with policy bodies including UNESCO, ICOM, ICOMOS, Council of Europe and national funders like the Arts and Humanities Research Council and Wellcome Trust.

Projects and Case Studies

Representative projects include large-scale OCR correction for the British Newspaper Archive and the Chronicling America corpus, a crowdsourced transcription program linked to the Zooniverse platform for ship logbooks from the National Maritime Museum and Royal Museums Greenwich, a visual similarity project across collections at the Tate Galleries and National Portrait Gallery, London, a linked-data reconciliation campaign populating Wikidata for objects from the Victoria and Albert Museum and British Museum, a geospatial historical atlas using maps from the Ordnance Survey and Royal Geographical Society, handwriting recognition applied to medieval manuscripts in the Bodleian Libraries and Cambridge University Library, and an oral history digitization pipeline for British Library Sounds and the BBC Archive. Case studies document reuse in exhibitions at the British Library and Tate Modern, scholarly publications in journals associated with the Royal Historical Society and presentation at conferences including Digital Humanities Conference, CHI Conference on Human Factors in Computing Systems and NeurIPS.

Impact and Reception

The Lab's work has influenced collection standards at the British Library, citation practices in digital scholarship at University College London and data-sharing policies at the National Archives (United Kingdom). Peer recognition includes invitations to present at the International Council on Archives and contributions to Europeana policy documents. Critiques from scholars at institutions such as the Courtauld Institute of Art and Institute of Historical Research have focused on algorithmic bias, provenance, and community engagement, prompting iterations in methodology and governance with input from Historic England, National Trust (United Kingdom), English Heritage and community archives.

Category:Digital humanities organizations