Paper 29

2018 Washington conference submission

Back to program.

From Digitisation to Discovery: IIIF Explorations with Edinburgh's Special Collections

Joe Marshall - The University of Edinburgh (United Kingdom), Scott Renton - The University of Edinburgh (United Kingdom)

Abstract: The University of Edinburgh has a large heritage collection, including art, museum objects, archives, rare books and manuscripts. The rare book and manuscript collections are substantial, with over 400,000 rare books, some 300 Western medieval manuscripts and 700 Oriental manuscripts. Up to now only around 2% of the collection has been digitised but we are now investing significantly in digitisation and have developed a Digitisation Strategy to prioritise work. The focus is on how we can unlock these unique resources to support the teaching and research of the University and the interests of the wider community, all in line with our open by default approach. From the RBMS collections we have identified as a priority for digitisation our iconics collection of star items such as the Celtic Psalter (probably the oldest surviving Scottish book still in Scotland).

However, we have identified that digitisation on its own is not sufficient to allow people to find and use this content. Particularly as regards the RBMS collections, there is a lack of good existing metadata and even when this is available the current image management systems do not draw on this effectively. Images are presented in isolation without contextual description, and consequently act more as decorative objects rather than as research resources. We felt that this collection lent itself to further experimentation with the IIIF model not just as a presentational tool, but as an approach to improve discoverability and usability.

Our image repository is very useful for image management, storage and "availability", but doesn't fit perfectly within our overall data/discovery solution. It is a IIIF compliant server though, and we have an existing model for creating logical, faceted, collections-based websites, supplied from CMS data. We aim to use that model to source this content.

The shelfmark is a logical means of grouping items to drive the generation of manifests, and is a good starting point (the Presentation API allows us to push beyond this). Because we are able to generate manifests dynamically, we can then push them into a repository which offers a persistent reference from other systems.

Further to this, we want the book-level data that we display to be correct, which means sourcing it from the library catalogue, and not simply picking up data from image records which are not subject to the same rigour. To this end, we are looking at making use of the ways to allow us to pull in library catalogue data, be it in MARC, RDF or JSON-LD form.

IIIF has other uses for us in this sphere: we are conscious that not everything digitised necessarily fits this model, and we can explain how IIIF can help with that content. We also realise it can help us identify cataloguing and digitisation gaps within the collections.

Finally, given the rich and varied nature of the Rare Books and Manuscripts we'd like to show that IIIF can meet the needs of complicated presentation material- the 75m-long Mahabharata Scroll- and has allowed us to capture translations and transcriptions.

Presentation type: 20 minute presentations (plus 5 mins questions)


  • IIIF and archival collections,
  • IIIF implementations from outside Europe and North America,
  • IIIF synergies with regional and national ongoing digital libraries, museums and archives initiatives,
  • Discovery of IIIF resources,
  • IIIF content communities (museums, manuscripts, newspapers, archival content, etc.)


  • rare books,
  • manuscripts,
  • presentation API,
  • manifests,
  • repositories,
  • scrolls,
  • translation,
  • transcription,
  • annotation,
  • discovery,
  • digitisation