Background
The National Galleries of Scotland’s Digital Engagement Strategy and Systems of Engagement project (2014 - 2018) focused on the improvement of digital access to its rich and varied collections. The previous incarnation of the website had been operating since 2007 and was focused on showcasing in-depth, highlight records from the collection. By 2014, only 6% of the collection was represented online (around 5,800 records), so both the platform and the publication processes had to be overhauled to bring the online presence up to audience expectations.
The nationalgalleries.org website in 2014
Systems/Architecture
The primary focus in this initial phase was to publish a data record for every artwork in the permanent collection, as well as long term loans. The development and initial launch phase of the platform coincided with a mass digitisation project that sought to produce a high quality image of every artwork in the collection, which would be catalogued in the Digital Asset Management System (DAMS) and linked to the relevant collection record in the Collections Management System.
As a manual import/export process would not be feasible at this scale, an automated ingest process designed by the SoE project manager is used to pull information from the relevant internal systems - using OAI-PMH for transfer of CMS records using LIDO-XML schema, and a RESTful API for transfer of images and data from DAMS, subject to the IPR and publication logic (i.e. the data is only ingested if the permissions are in place).
nationalgalleries.org website architecture. Image credit: Ashley Beamer
The data is then combined into a single record in a Drupal install that acts as a middleware tier; where content creators can add supplementary data as required without altering the core collection records themselves. In addition, custom taxonomies can also be created for data publication (such as the art glossary and subject tags).
Data from middleware is indexed in Apache Solr for fast, responsive search and discovery; derivative and tiled images are also generated and stored in Amazon S3 for delivery to the front-end.
Front End
When the user visits the collection online, they are able to dive in to explore all the artwork records, or can be led in through a set of curated entry points administered using the custom taxonomies, such as subjects, highlighted artists or curated collections of artworks.
As records were now indexed, this allowed for expanded search and filter capabilities, including dates, artwork locations, artwork types and more:
The artwork record pages themselves were redesigned to present an expanded set of tombstone data, along with additional content such as expanded descriptions and media content, and any relationships defined within middleware. These were all linked back into search to allow users to continue their exploration of the collection.
With the new mechanisms for image delivery we were also able to provide an image download service (subject to rights) as well as a zoom facility powered by the OpenSeaDragon viewer.
Data Enhancement Projects
With the publication of ~92,000 records, it was vital that we provided visitors with accessible routes into the collection. Taking existing taxonomies from the collections management system, we constructed and enhanced these in middleware.
Subjects allow users to browse artworks through the subject matter they contain, concepts they relate to in social history and underlying systems and structure behind the artworks.
Glossary terms allow users to explore common art history terminology such as styles, movements, techniques, and materials. These taxonomies are supplemented with introductory content, media and related features to provide an explanation of the term, along with the related artworks and artists in the collection.
Artist pages presents enhanced content on around 1000 of the ~14,000 artists recorded in our collections management system. This includes biographies produced by the galleries, related content and data from across the organisation, including exhibitions where work by the artist is displayed.
Where supplementary biographical data is accessible from services such as Getty ULAN and Wikidata these are pulled through dynamically. This was achieved through a project that reconciled the artist records in our CMS with the respective Getty/Wiki records. The identifiers are then stored in the galleries artist records and the data is pulled through to the page automatically using the respective API’s for each service.
Finally, curated collections are a taxonomy that allows gallery staff to create showcases of artworks around a given theme for users to explore. Examples of how these can be used include recent acquisitions, highlights of particular collections, or they can be used to hone in on a particular strength of the collections e.g Scottish contemporary art.
Results
By 2017-18, over 90,000 artwork records had been published online, many for the first time. Of these, around 60,000 were published with images - a number that continues to grow as the galleries digitisation programme progresses. Supplemented by over 100 glossary entries, and enhanced data on over 1000 artists, audiences are now able to discover and explore every artwork in the collection.