Annual report 2022/Story: Improving search and useability – Continuous data uploads and improvements

Från Wikimedia
Hoppa till navigering Hoppa till sök

In this Story we are focusing our continued work to make images from the Wiki Loves Monuments competitions easier to find and reuse with help from Structured Data on Commons. Enriching the photos with data in a structured format both boosts the efforts of the volunteer photographers and makes the world's cultural heritage more visible.

Structured Data on Commons makes it easier to find photos of protected cultural heritage around the world. This map shows photos from Wiki Loves Monuments in Ukraine that have both geographical coordinates and depicts statements.

Wikimedia Sweden's work with Structured Data on Wikimedia Commons continues. After last year's efforts, we have developed good working methods and routines for processing large numbers of images. It has grown to be a significant part of our long-term strategic work with the world's cultural heritage on the Wikimedia platforms. Throughout 2022, we have been working with images from the world's largest photo competition, Wiki Loves Monuments, to make them easier to find and reuse.

Adding structured data makes it easier to answer questions like: Which country uploaded the most photos to the competition in the last five years? In which regions were the most photos taken? Which protected objects – such as archaeological sites, churches and other buildings – lack photos, thus giving our volunteer photographers a great opportunity to contribute to the Wikimedia platforms? In cases where the images are provided with geographic coordinates, which is often the case these days when our cellphones can take fantastic photos and automatically geocode them, it's also possible to add the coordinates as structured data and see them laid out on a map.

Our goal in 2022 was to add 1 million structured data statements – that is, to which regional competition the photo was submitted, where it was taken and what it depicts. Of course, we did not do it by hand, but processed the information that the photographers themselves provided in the description text. Because of that, it's not possible to identify all the depicted objects automatically. It requires that, for example, the church in the photo has an identifier, such as (in Sweden) a code in the Building Register, which in turn can be linked to a corresponding Wikidata item. Fortunately, many countries have good coverage of their heritage sites on Wikidata, which is an excellent springboard to automatically process the image descriptions and convert them into structured data.

What about our goal? We achieved it by adding 1,290,000 statements to 724,000 photographs from 20 countries. The largest image collection came from Ukraine with over 300,000 processed images. Prioritizing this country was an obvious choice at a time when the world's eyes are on it, and its amazing cultural heritage is actively threatened.

The war has not only claimed lives but also destroyed places of cultural, historical and religious significance to the nation, which is why it's especially important to make the vast collection of images – Ukrainian photographers have always been diligent participants in Wiki Loves Monuments – more accessible to both the Wikimedia community and the public.

When Wiki Loves Monuments took place in Suriname for the first time, over 500 photos were uploaded. In the winning photo we can see the building of the Ministry of Natural Resources.

With the help of SDC, it is now easier to find beautiful images of cultural heritage from both large and small countries – everything from Ukraine to Latvia and Norway – that can be used to illustrate Wikipedia articles. A real highlight during the year were the pictures from Suriname. It was the first time Wiki Loves Monuments was organized in this South American country. We partnered up with the competition organizers to convert their Wikipedia-based lists of the country's cultural heritage to Wikidata items, which, as previously mentioned, is a prerequisite for being able to tag the photos with the correct depicted object.

Wiki Loves Monuments is held around the world every year, and we are happy over having developed a workflow to handle new local contests and their amazing photos. Seen in a larger perspective, the work we and the community are doing is an important piece of the puzzle in strengthening Wikidata's position as a global hub for cultural heritage data. Each image linked to the depicted object makes Wikimedia's linked data collection richer and more interesting.