By Martin Poulter, Wikimedian in Residence at the Bodleian Libraries, Oxford
The speed at which Wikidata is acquiring descriptions of paintings, sculptures and other museum holdings is impressive, but there is much further to go. It’s ironic that at the same time, we already have an enormous art database hiding in plain sight.
The Commons index of Indian art may be the largest digital collection of Indian art ever created. Its collection of Cities of France in art may be the biggest such index in the world. No institution has collected as many pictures of astrolabes as are in the Commons astrolabe category. The superlatives go on.
The problem is that what we have on Commons isn’t yet structured data: it’s not possible to get all images matching a chosen set of criteria; just the criteria that are pre-baked into the category system. Meanwhile, there are items from cultural institutions which are described in Wikidata and also have an image in Commons, but the two are not yet linked up. It’s crazy but it’s just the result of the order in which our platforms developed.
Structured data is coming to Commons in the longer term, but there are things we can all start right now. We can use the existing structured database – Wikidata – to improve the content and findability of photos on Commons.
The secret weapon is the Art Photo template:
- It imports data about the artwork from Wikidata, just needing the relevant Wikidata identifier (a Q followed by some digits)
- It keeps distinct the properties of the photograph and of the object in the photograph.
- It has fields for photo date and photographer, distinct from the date and authorship of the item. If you take a photograph in 2019 of a statue from the 9th century, we want to avoid the ambiguity of a single date field..
The resulting entries are more multilingual, more detailed, have links to further information, and will automatically draw updates from Wikidata. Unit conversions (e.g. between inches and centimeters) are done automatically. In other words, we take some dull work from human beings and make the computer do it instead. All this while decreasing the amount of wiki-code on Commons!
If you’ve photographed a museum exhibit or piece of public art and shared the image on Commons, or if you’re interested in art works of a certain kind, I urge you to take a look through those photos and to search Wikidata to see if it describes what you photographed.
If the item does not yet exist on Wikidata, it’s surprisingly easy to add it. Take the Wikidata tours to learn about the interface.