L.A. Ring, A Landscape near Bryrup, Jutland, 1888 (on the map).

In March 2021 we placed 4.000 SMK artworks on a map of Denmark. We did it by combining existing data with machine learning and the help of kind human beings. And we did it by leveraging the SMK API which will ultimately ingest the refined location data to allow future re-use.

The art map of Denmark at danmarkskort.open.smk.dk

That, dear reader, was the abridged edition. For the hows, the whys and the what nows, please scroll on!

The agony of size

The SMK collection holds some 260.000 artworks. Inspiring? Sure, but also potentially terrifying. Where to begin? How to navigate? As a regular Danish person, you may have no obvious point of reference.

Enter: The map. A map makes things decodable and very personal. You were born somewhere specific. You grew up somewhere specific. You live somewhere specific. A painting portraying one of these sites is immediately identifiable. It’s something you may know very well indeed.

Now, seeing a well-known place rendered through the eyes of art may induce memories. But it may also inspire you to see a landscape afresh. And in the time of COVID, it may let you travel virtually or supply new inspiration for your physical weekend trip.

Putting paintings in their place

Having established the virtues of cartography, how to proceed? SMK’s collection is registered in the national Danish collections database SARA, an acronym referring to the prosaic “collections registration and management”. Some artworks have rudimentary location data about what they depict such as “Northern Zealand” or “Copenhagen”. Often not very precise, but a place to start!

John Christensen, Folkeliv på Kapelvej, 1932 (see on map).

To narrow things down we analysed artwork titles using a Named Entity Extraction technique with a Multilingual Bert model*.

Machines were tasked with teasing out place names which, combined with the museum keywords, were run through Google Maps API and the artworks were placed on a Google map. For instance, John Christensen’s Folkeliv på Kapelvej (“Street life in Kapelvej”) was now placed, not just in the area Nørrebro in Copenhagen, but on the actual street named Kapelvej.

Importantly, we limited the results to Denmark (including Greenland and the Faroe Islands) and excluded certain location types such as cafes (which are often named after places, bound to cause confusion).

Vilhelm Hammershøi, Landscape on the Island of Falster, 1890–1891 (see on map).

The level of precision achieved? Useful, sometimes spot-on, but not always convincing. We found close to 4.000 places but often the location was very general.

For Vilhelm Hammershøi’s Landscape on the Island of Falster all we could find was “Falster” and so the painting was placed in the center of that fine island even though it actually seems to be painted by the sea.

Now, obviously the dataset wasn’t extremely daunting and we could have refined the locations manually ourselves. But while common sense might improve precision somewhat we are hardly experts on every specific hillside, park bench or cliffside. But out there, someone is bound to be! So we published the map and asked the public for help.

Artificial intelligence meets local expertise

On 17 March 2021 we published the map. At the same time we started up a dedicated Facebook group and sent out a general press release. In the Facebook group we invited members to join us on a trip across the country from East to West and as this trip (consisting of posts of highlights from the area) progressed we contacted local news outlets and relevant Facebook pages or groups. We emphasised two attractions of the map: See your hometown through the eyes of art (for the general public) and Help us improve the map (for the more dedicated user).

Together, these efforts stirred up considerable attention. A range of local media picked up the story, and as of 9 April 2021 we’ve had 16.000 visits. Even more interesting, I think, is the fact that our geographical reach is far more evenly distributed than we’re used to. We often tilt strongly towards Copenhagen but this time only 35% of users are in the capital (the home of 23% of the population).

The distribution of users (main hot spots)

Initial feedback was quite overwhelming. Also, it was chaotic. We’d — naively! — planned to manage dialogue either through the built-in features of the map interface or the dedicated Facebook group. A nice plan, but immediately emails started pouring in in the form of social media comments, DMs to our private accounts, emails to the museum’s main address and even to our director.

Sure, there were concrete suggestions for moving map pins. But also, of course, there were questions about digitisation practice, image resolution, requests for clarification and (sometimes kind, sometimes less kind) suggestions that we read up on geography. Honestly, they were all extremely welcome and I only hope that not too many fell through the digital cracks.

Many commented that locations with identical names had been mixed up. Many also commented that Google had sometimes guessed a little liberally when places no longer existed (“Sydstrup” no longer exists, so Google had guessed “Skrydstrup”). But perhaps the most interesting comments were ones that helped us clean up errors in our core data. Apparently, when transcribing venerable old paper cards, some museum worker — at some point in time — had assumed that “Femø” (south of Zealand) should be “Fanø” (west of Jutland). And somehow we’d typed in “Kalvø” instead of “Kalø” in the title of Janus La Cour’s A Showery Landscape. Kalø Vig, Jutland placing it around 160 kilometers too far South. We’ve essentially (and rather accidentally) crowdsourced the proof-reading of our metadata, and I do think this qualifies as something of an eye-opener.

Where do you want to travel next?

We’re continuing our journey Westwards (and learning tons about Danish geography as we go along). When we’ve reached absolute (Danish) west, we’ll work on importing all the new coordinates into our API, and perhaps also into the collections database itself.

Meanwhile, many have asked why on earth other museum collections aren’t on the map.

Not surprisingly, people don’t really care which museum semi-arbitrarily stores which artworks — they care about art. Including other collections is way beyond the scope of our project, but the logic is totally sound. In Denmark, we’ve recently struggled to merge most museum collections databases into one and while I realise that this task was momentous in itself, of course the next step should be linking up the data in new and useful ways. We hope our little map may provide some inspiration for how to work with the nation’s museum data in the future.

Today we're launching 'Kunstens Danmarkskort' (The Art Map of Denmark) – placing @smkmuseum artworks on a Google Map 🍾 The map: https://t.co/gHXNfKUdnE (interface in Danish)#smkopen #openglam #dkmuseum pic.twitter.com/sZTtut29fM

— Jonas Heide Smith (@jonassmith) March 17, 2021

• I’ll be happy to pass on any technical questions you may have…

The team

• Christina Jensen
• Nikolaj Erichsen
• Merete Sanderhoff
• Jonas Heide Smith
• Michala Rosendahl

With very special thanks to Sofie Glargaard, Kim Brasen, Tanja Larsen and others.

The signs are all around us: Museums are abandoning the museum-as-building paradigm and even stepping beyond the museum-as-building+website “tiny addendum model” to embrace the idea of the museum as platform. Not in a coordinated rush, and certainly not at similar speeds but as an audible rumble fueled by the affordances of digital technologies.

If I may very briefly summarize, the idea is this: The interface to our collections is now a myriad different views through a myriad different screens over which we have very little control. The museum, in other words, has become increasingly distributed (see also Bautista & Balsamo, 2011) and strictly prioritizing one interface over another is non-trivial. Or as Nancy Proctor exemplarily notes “a bricks-and-mortar museum is an analog platform” (Proctor, 2010).

Now, while this new model may be increasingly accepted in some variation it often remains largely metaphorical. While digitizing one’s collection, surrendering control, and making files accessible through third party platforms is brave, commendable, and challenging in itself it does not fulfil the digital potential of most museums. Let me explain why I believe we need to be more reductionist yet more disorganized in order to succeed.

Towards a thick connection

You may have accepted that the internet is the prime repository for shared knowledge (or ‘information’ if you insist) in our societies. For any knowledge institution we are long past the point where choosing not to have a strong impact online is the choice in need of an explanation.

Many institutions have material online in some form and thus have at least some connection to the web. But the usefulness of these connections runs the gamut. A “thin” connection is one where very little material is published and/or where this material is unorganized. It’s what you get if you, say, place a collection of photos for download on your website. The very dedicated, the very knowledgable, or the very lucky can find them and use them but the downloaded photo will typically have very little associated information. Your museum may know a lot about the object but the user will be able to find out very little. In a sense, then, your collection will be connected to the shared knowledge bank though a very thin connection.

At the other end of the spectrum, a “thick” connection is one where the museum’s knowledge resources are published far more systematically than the occasional JPG. It’s one where material is both accessible and useful to individuals and to systems. And it’s one where ideally all of the museum’s relevant knowledge comes packaged with the object. It’s also one that requires a leap into reductionism and disorder.

The case for reductionism

To the computer on your desk and the phone in your pocket, everything is data. Your device can run mind-numbing accounting software, play Handel’s Music for the Royal Fireworks, or display your flight schedule and in an important sense it will do it all equally well. It won’t care about specifics — it’s a platform, and a wildly important one due to rampant reductionism.

Think of the standard shipping container (that probably contained your device at one point) — a similar mechanism. It’s context-free blandness is its very strength: It’s highly flexible and can pretty much contain everything.

Don’t worry too much what’s inside as long as they stack well. Digital material is somewhat like a shipping container.

Museums have digital content. But in order to make this content flexible enough to be truly useful we need to see not videos, research articles, blog entries, nor exhibition websites but content blocks. In a sense, then, we need to do what we are trained not to do: We need to ignore context.

Does this sound disturbing? If so, the notion may have made you think of the disreputable practice of cross-platform publishing; the (often problematic) idea that the same content can work well in many different formats. But the reductionism we’re looking for here is not this, but rather that of a library: To be useful, a library defines any book as “simply” an instance of books, thus ignoring typography, quality of writing, the mental state of the author, the imagined reactions of the reader etc. But, emphatically, the library does not reformat books on purchase to look the same and museums have no good reason to do this either.

In other words: We need to forget the properties of our materials that are irrelevant to organization without, of course, destroying these properties.

Nicolai Abildgaard’s ‘The Wounded Philoctetes’ from 1775. Artwork placed by SMK in the public domain. Technically available but hard to find and use without specified relationships to other material.

Order through disorder

Which is the best way to organize our digital content blocks? If you’ve ever tried to draw up a sitemap for a complex website you know that this is dangerous territory. But in an important sense the answer is “in every way”. Organizing digital content is decidedly not the same as organizing physical objects. A physical object can be in one place at a time. In a physical art museum, you can organize your exhibition chronologically, by theme, by style, or by painter but you can only choose one. On the museum’s website, while you cannot have the physical artworks themselves, you can have all your representations and all your material in every way at once and this requires a footnote: We need to leave behind the shipping container metaphor — despite all their flexibility shipping containers can only be in one place at a time. Let’s refer to our material instead as elements.

A sitemap has its uses but it’s a strangely physical exercise as it imposes a particular hierarchical structure on elements that really don’t need it and will necessarily correspond to the mental model of only very few users (on deeper levels it’s likely that no one user will find the structure fully logical). What we need is a way to let users choose their own structure based on their own definition of relevance. And for that, we need to establish connections between our elements by introducing what we could call associators. Associators are labels used to form relationships, in other words they are ‘metadata’.

Of course, the more common name for such metadata is “tag” or even “keyword” but those may give the wrong impression as an associator is better thought of as anything that defines a relationship between elements. On this view, we can identify at least four types of associators:

• Organic. Keywords contributed by someone, whether museum staff or uses.
• Machine-based. Keywords contributed by computer analysis of image content (for instance)
• Found. Properties of the file itself such as camera metadata, document length, color distribution of an image.
• Implied. Relations gleaned from user behaviour. For instance, a relationship can be established between two objects that are often seen by the same user.

Using a scheme such as this, rich relationships can be built between objects in a collection. No all-encompassing authoritative taxonomy is needed. The price, of course, is a certain unpredictability. But the advantages are legion as you (and your users) will now have a much more versatile way of organizing everything based on any requirement.

Towards thicker connections

To successfully be a platform — in the very concrete and somewhat technical sense described here — museums must not just publish their material online but meaningfully connect their materials to each other and to the web. I’ve left out the technical specifications on purpose since the whole point here is to think beyond concrete platforms but in outline, it’s a model we’ll be pursuing at SMK — The National Gallery of Denmark as part of the SMK Open project in the years to come. The idea of the distributed museum is by no means new but if we can establish much more powerful “thick” connections between museums and the web we may just be pushing our relevance into a whole different league.

This article appeared in Museum iD, issue 21 (August 2017)

Bibliography

Proctor, Nancy. “Digital: Museum As Platform, Curator As Champion, in the Age of Social Media.” Curator: The Museum Journal 53, no. 1 (2010): 35–43.

Bautista, S, and Anne Balsamo. “Understanding the Distributed Museum: Mapping the Spaces of Museology in Contemporary Culture.” Museums and the Web 2011. 2011. <http://www.museumsandtheweb.com/mw2011/papers/understandingthedistributedmuseummappingt.html.>_

My presentation on #openglam, @smkmuseum, and the possible future(s) of #musetech from #museumideas today https://t.co/0dDZhpcGwr

— Jonas Heide Smith (@jonassmith) October 5, 2017