From the British Library to Flickr to Europeana - 60000 public domain images on the move

This blog was written by Henning Scholz - Partners & Operations manager at Europeana and Tom Miles - project manager at the British Library.

In December 2013, the British Library released over one million public domainimages onto Flickr Commons. 60,000 of these images are now available in Europeana for your creative re-use.

The images were taken from 17th, 18th and 19th century books digitised by Microsoft and generously gifted to the British Library. This vast release of material was carried out by the British Library’s ‘BL Labs’, funded by the Mellon Foundation.

The metadata of the scanned images was sparse - only the title and author of the book were available. Fortunately, Flickr users began tagging the images and by November 2014, more than 100,000 contextual tags had been added to more than 60,000 images. For example, an illustration of a bird like the one below might be tagged with ‘fauna’ and the name of the species.

Lesser White-fronted Goose (Anser erythropus) from ‘A Spring and Summer in Lapland: with notes on the fauna of Luleä Lapmark’ (by H.W. Wheelwright) - Public Domain marked.

The high resolution scanned images are wide-ranging in content, including maps, illustrations, decorations, landscapes, wall paintings, portraits and more. And as they’re in the pubilc domain, they’re perfect for creative re-use. That’s why the Europeana Creative project wanted to make all the well-tagged images available in Europeana.

The ingestion of these images into Europeana is great news for our creative industry partners. The pilot application for the ‘Design’ strand of the Europeana Creative project is ‘Culture Cam’, a webcam-based similarity search tool offering designers, artists and all creative people the ability to explore and access Europeana images in a fun, playful and intuitive way. The British Library Flickr images will be integrated into the application within the next few weeks, making them available for exploration.

So, how did we do it?

The British Library prepared the metadata for the images in close collaboration with Europeana and The European Library - the first time Flickr images have been prepared for Europeana. One of the challenges was to ensure that the image links all point to the right place on Flickr. In addition, each metadata record needed to link to the full pdf of the book and the British Library book viewer, so that people can access the book and either download it or read it online, seeing the image in its original context.

Another challenge was to make the titles for every object in Europeana as meaningful as possible. At first, we used the first Flickr tag of each image as the Europeana title but many items had the same tag, e.g. thousands of images were tagged ‘map’. So we added the title of the book to add more context. The image above is a good example. The tag from Flickr is ‘Lesser White-fronted Goose (Anser erythropus)‘ while the book title ‘A Spring and Summer in Lapland: with notes on the fauna of Luleä Lapmark (by H.W. Wheelwright)’ provides the necessary context.

It was important to visualise how this data would display on Europeana. Would just one word be sufficient as a title? Could the title be improved by globally combining one field with another? At what point should all this global editing take place? Close collaboration between Europeana, the aggregators and data providers is currently the best way of making the right decisions. Meanwhile, Europeana is working on streamlining and improving the aggregation infrastructure and ingestion workflows to make it easier for institutions to participate and share their best data.

The British Library Flickr photostream has demonstrated how useful ‘crowd-sourced’ activity can be in enhancing records. On Flickr, many individuals adding tags unknowingly worked as a team. For example, some individuals simply tagged images with the word ‘map’. Others then added geo-references to those images. One individual might tag 100 images as ‘bird’, allowing a more expert user to add more specific information. Crowd-sourcing should not be seen as a replacement to professional cataloguing, rather, a tool in the curator’s workshop to be used with skill and judgement.

Get involved

This work demonstrates a new way of creating attractive and re-usable high quality content for Europeana. It also solves a known issue for memory institutions that want to share their precious photo collections with Europeana while hosting them on Flickr. We welcome more data partners to use this approach to feature and share their best material. For any questions, please email Henning Scholz. For any questions about the British Library’s role in Europeana Creative, please email Tom Miles.