Strategy and open data at Wellcome Collection

London’s Wellcome Collection is a free museum and library exploring health and human experience. Inspired by the collections assembled by Henry Wellcome, it encourages great ideas about health by connecting science, medicine, life and art.

Last year, Wellcome Collection set out on its new digital strategy and began work on improving access to its collections. Starting with its image collection, they migrated the content from the old Wellcome Images site to a new platform, to form part of the Wellcome Collection website. Tom Scott, Head of Digital Engagement, shares his thoughts with us.

Last October, you published a digital strategy paper for Wellcome Collection. Six months on, how are things progressing?

We’re making good progress. Last quarter, on wellcomecollection.org we tackled What’s On, exhibitions and events. User research helped us identify how to help our visitors, it helped us identify what they wanted practically and emotionally. Usability testing then helped us understand whether our design ideas met or exceeded those expectations and needs. The result, a combination of improved interface design and a new approach to content and imagery, has significantly improved the user experience -- in the last quarter we saw a 15% increase in users and 135% increase in ‘reading time’ (active time on content pages).

Behind the scenes, we’ve also done a huge amount of work to bring our catalogue across into the new platform. In addition to the c. 110k images we migrated across from the old Wellcome Images website, we’re now exposing 1.16 million catalogue records via the API and as downloadable datasets.

Because we will also be merging in the archive catalogue (ISADg records), intra item metadata (METS and ALTO data) and enriching the data through machine learning (e.g. to identify dates and other entities), the new platform restructures the data into a single, coherent domain model. On the front-end, this is important because it means, for example, that you won’t need to understand how an item has been catalogued before you search for it.

There’s still work to do on the API and the underlying data before we start building a website with it but, if you are interested, we’ve published a preview of this data, details are all available at http://developers.wellcomecollection.org 

Open access to Wellcome technologies like Cardigan, APIs and source code seems to be central to your strategy. Why is this important and what engagement have you achieved externally?

Being open is in our DNA. Wellcome exists to improve health for everyone by helping great ideas to thrive. For ideas to thrive, we believe, they need to exist in the open so that others can integrate them, use them, learn from them and improve upon them. We hope that is true for the data, code and content we are publishing as part of our work.

But it works both ways. Under the hood we use Loris as our IIIF image server. Loris was an existing Open Source IIIF Image Server and in adopting it we have benefitted from all the work that had gone before, but we’ve also improved Loris and that benefits all the other Loris users.

Working in the open also helps us to do a better job of building a scalable, maintainable service. For example, we use the public catalogue and IIIF APIs to build wellcomecollection.org. We don’t have a secret backdoor or private API, we use the same public APIs available to everyone else, and that is a good thing because it helps us enforce a separation of concerns.

The platform team (the team responsible for the API) care deeply about how the data is gathered, modelled and stored. The experience team (who are responsible for the website UX) have no vested interest in how gathering, modelling and storage happens, as long as it happens. On the other hand, the experience team have their own concerns about what they are receiving, how to render it and how users interact with it.

Obviously, we could enforce this separation without making everything open and public but in doing so we have no choice and that helps us stay honest when deadlines loom!  

Wellcome Collection makes over 100,000 images available to download under a CC BY rights statement. Why was CC BY chosen and what was the decision-making process?

We have a few more than that! We are digitising our 19th century book collection, archives, manuscripts, recipe books and much more. To date, we’ve digitised and made available (at wellcomelibrary.org) around 36 million images (a number that is rising at around 250k images a month), representing around 250,000 items. This is in addition to the 100,000 ‘visual images’ originally brought over from Wellcome Images. Over the coming months, we will migrate the wellcomelibrary.org images across to the new platform too.

Most of those images are licensed under CC-BY - some aren’t - mostly where we don’t own the copyright. We selected CC-BY because we wanted to make it as easy as possible for researchers and others to use and reuse the material and because we wanted to reach as many people as possible, not just on our own website but also elsewhere - on Europeana Collections, Wiki Commons, the Internet Archive and elsewhere.

How has the new site been received?

We run weekly user research and usability testing, online polls and A/B tests and, based on the insight we’ve gained through that work and web analytics, we’ve got a pretty good idea where we’ve got it right and where we need to work harder. On balance we are definitely ‘winning’. Usage and engagement are both up and increasing and qualitative feedback is positive. The teams responsible - product design and development, photography, editorial content and production - have done an amazing job and I am in awe of what they’ve achieved.

Why is it important for the Wellcome to publish its data on Europeana Collections?

Europeana’s mission to help build on Europe’s cultural heritage and to contribute to an open, knowledgeable and creative society is something we want to support, it is something we believe is important too. I hope that by working together we can help contribute to a more open, creative society where ideas can thrive.