Every three weeks or so GLAMtech links, September 2 2018

The eagle-eyed among you may have noticed I did not publish my listicle of links last weekend. The weather was just too fine and so I did some a day hike and some archaeological exploration. But fear not! With one week’s delay here are 10 links to interesting GLAMtech reads.
#1 SketchFab launches support for very large datasets and models
This new feature is in alpha and should prove useful to e.g. archaeologists who work at the “landscape level”.
#2 The National Gallery of Denmark releases new website
Jonas Heide Smith shares the technical design intent behind the recently launched, and very beautiful, renewed website of the National Gallery of Denmark. A key design features is the choice to opt for a headless CMS (in this case WordPress) and using a DAM with a IIIF-endpoint and develop the website itself as a client of those. Technobabble!?
The point is that this API-centric approach is that it will give the museum greater flexibility in designing and developing futur user experiences, will make it possible to allow 3rd party developers to use those same APIs to also create new applications and experiences, and will make it easier and less costly in the long-term to maintain their tech stack (because it’s made up by multiple modules that can be maintained and replaced separately, not a monolith that at some point hast to be either kept in all or all thrown away ).
#3 2018 Horizon Report on #EdTech
For professional reasons I have a renewed interest in EdTech, esp. Open Education Resources. EdTech is a business that, IMO, is beset by a plethora of companies that value their bottom lines much more than students receiving a good education (or the chance to educate themselves). Still, I will read this with my semi-open mindset that technology can make a good teacher better, but with the fear that it can make a poor teacher a worse one. Fair and balanced?
(This tweet doesn’t seem to want to be embedded. If the Twitter card doesn’t show for you just go here to read the report.)
#4 Nationalmuseum launches a (IIIF-powered) virtual exhibition
Did I mention in link #2 above that APIs and API-centric system architectures can make it simpler for museums to create new user experiences? Yeah, I did.
#6 An introduction to machine learning from a library perspective
A subject we all do not need to be experts in, but should have a basic understanding of. Why? Because machine learning based software is already changing the way we produce, publish and consume content, incl. heritage content.
#7 An innovation process and the apropriate methods in each phase
Like EdTech above Innovation Management is something I have to learn about rather than something that comes naturally to me. I think there are a lot of GLAMs who struggle with digital transformation and the interplay between changing the way we work and the “corporate culture” that both results from and shapes that way we work.
At my job we’re also in the early stages of creating am incubator for companies that want to grow by creating digital products and services for GLAMs and so this is a process we ourselves need to be able to guide them through.
#8 Why do museums no longer create games?
There was a time when many museums created (eduational) games. Or at least talked about creating them… Why is there less action and talk of it now? This Twitter thread provides some answers.
#9 Europeana looking for test users
Do you use Europeana Collections? Or did once but no longer? My former colleagues at the Europeana Foundation is looking for insight into users’ motivations and pain points.
#10 Machine learning applied to the V&A’s photography collections
Did I mentione above that machine learning is already impacting the way we create, publish and consume content? Yeah, I did. Here’s an example from Olivia Vane where she’s using machine learning to power a discovery experience based on image similarity.
The first early adopters in the GLAM-sector are now beginning to apply machine learning. I think already 5 years or so from now cataloguing image collections will be AI-assisted — the cataloguer will be able to choose to accept and index (or reject) what the vision API that comes built-in to their CMS or DAM suggests for each image.
FIN