We were glad to see this recent tweet from Andy Mabbett:

New Mix’n’Match catalogue, for importing 24.5K @WhatDoTheyKnow IDs (and thus links) into @Wikidata as part of the web of #LinkedData.

Just started, so automated matching not yet completed, but you can help by manually matching the rest.#OpenData #FoIhttps://t.co/mxINch1llo

— Andy Mabbett (@pigsonthewing) May 5, 2020

Andy has imported the IDs of every authority listed on our FOI site WhatDoTheyKnow into Mix’n’match, a tool for helping to link a dataset with existing Wikidata entities. Once a match has been made, the URL of the body’s WhatDoTheyKnow page is available as one of its identifiers (specifically, P8167).

This means that anyone running a project that utilises Wikidata will have the option to include WhatDoTheyKnow data in their site or app.

Andy says, “Wikidata acts as a hub for all sorts of databases and identifier systems. For example, it can be the only way of linking (programmatically, in the linked data sense) an MP’s official parliamentary record to their IMDb entry. I do a lot of work making that happen. As a regular and satisfied user of WhatDoTheyKnow, it appealed to me to add that site’s 24.5K listings of UK public bodies to the mix.”

The best-known site relying on Wikidata is of course Wikipedia, so in theory it would now be feasible, say, to include a template that automatically pulled the relevant WhatDoTheyKnow link into Wikipedia articles about authorities, or to build a browser extension that provided those links when the user visited such articles.

It would also be possible for us to pull information back the other way, so for example we might consider importing the first paragraph of a Wikipedia page for a body and using it within the introduction, as a way of providing context.

The matching of WhatDoTheyKnow authorities confirms which Wikidata URI (Uniform Resource Identifier) relates to each, meaning that these can now be used in “sameAs” metadata headers, scehma.org markup, etc. We think this might have a beneficial effect on the way search engines treat our pages in the future — something we’ll be keeping an eye on to check if that’s true.

Additionally, this works as a nice proof of concept that we can potentially recommend to other Alaveteli sites around the world, given that the Wikidata project is, of course, international.

But first, the bodies need to be checked with the Mix’n’match tool. At the time of writing, 1,302 bodies have been resolved, and can be seen here. Anyone is welcome to help by confirming more matches: just log in with a Wikimedia account.

Thanks to Andy for this initiative — it’s great to see the potential of our data being widened in one fell swoop.

There has already been a mutual benefit to this linking. WhatDoTheyKnow volunteer Matt has been able to use examples of failed matches to find cases where our database needed to be brought up to date with name changes. At the same time, Andy says it has helped him and his fellow Wikidata volunteers to create new items about councils and other bodies that were in WhatDoTheyKnow but not Wikidata.

Richard, also one of WhatDoTheyKnow’s volunteer team, says, “I’ve often thought there’s a lot of overlap between what we do on WhatDoTheyKnow and what Wikipedia volunteers are doing — we’re both maintaining lists of public bodies — so any tools for closer collaboration are great.”

—
Image: Carl Nenzen Loven