ontologies

2022 - Week 17

Bot farming

Our regular reader will of course be aware of - and indeed following, one hopes - both made-n-laid and tweatytwacker, our earlier dabblings in Twitter automation. This week our family of bots expanded to incorporate five new accounts that tweet whenever an answering body provides a written answer - or a correction to an earlier answer - to a parliamentary question:

A thread from Anya explains more and comes complete with a call for interest in other answering bodies our readers might like to see botified. David was the first - and indeed only - reader to join the queue, which means, by late Friday we’d added @DFEAnswers for all your educational needs.

As your correspondents type, @FCDOAnswers has already gained 67 followers. Including at least two APPGs and a small number of people who seem to work for the FCDO. Which is not bad for a set of accounts launched in a pub of a Tuesday evening.

Not content with syndicating to Twitter - and with his finger poised, as ever, on the pulse of the zeitgeist - young Robert has jumped in to further syndicate from Twitter to Mastodon. How modern are we?

Proro-horror-gation

No sooner did we emerge refreshed and reinvigorated from Easter recess than Parliament prorogued. Or rather, Parliament was prorogued. As has become tradition, Librarian Jayne took this as her opportunity to publish our own take on statutory instrument related sessional returns. Again in the form of a tweet thread. A tweet thread that proved quite ridiculously popular. We do not kid ourselves that we’re working with mass market material here, so a tweet thread about SI statistics gaining 13 retweets and 50 - count ‘em - likes, is not to be sniffed at. By late Friday the followers count for made-n-laid had jumped by 150 and passed the 500 mark. This to include at least one Whips Office, a handful of Members, a scattering of Members’ staff, a collection of clerks, a profusion of professors and a quorum of QCs. Now, anyone that knows us will know that we are not stats and graphs type people. All efforts at measurement tending toward the what can be measured, because what’s useful to measure is usually too difficult. Anything of any real value usually being unmeasurable. Nevertheless, 500+ followers for an account tweeting SIs does feel quite remarkable and strengthens Jayne’s resolve to continue on her quest to be the world’s shyest social media superstar. She has threatened to quit if the number tops 1000. So please don’t all rush in. Seriously.

Prorogation is not just an opportunity to recap though. It also brings work. But you know us. Never shy of a decent day’s graft. Which is why Tuesday found Jayne, Robert and Michael hunched over pixels and prodding at database tables to make our beloved egg timer of parliamentary procedure prorogation proof. Which left poor Jayne lumbered with the thankless task of updating end dates for instruments before Parliament. It’s not all glamour and Instagram. The sooner we can couple procedural work packages to the egg timing thereof, the better for all concerned.

You’ve just given me your last Standing Orders

A wee while back we struck gold when boss brarian Bryn managed to conjure up cash to turn our standing order prototype into something more editable. This week, Anya, Jayne and Michael got their first glimpse of the new system, joining James for a pixel-based demo. And very good it was too. Very good. Really, really pleasing. James has taken Michael’s shonky Ruby code, sped it up quite considerably, added user accounts and made all the tables editable. And not just this. You can now take a revision set, press a ‘clone’ button and the whole set of standing orders and every fragment in them gets copied into a whole new tree, and in less time than it took Michael’s code to render a list. We’re not yet in a position where we can show you work in progress. And even when we are, you’ll miss the nifty edit and clone stuff but, needless to say we are pleased.

So pleased in fact, that Michael has since prodded Rad and Tom from the research team that first donated the standing order data and asked if they might want to use the same system for their work in other legislatures. Tom has kindly offered to sense check the editing interface. And given he has more experience in compiling this data than anyone, that would definitely not be time wasted.

Expanding our horizons

Our regular reader may well be under the impression that our crack team of librarians spend all day churning out data models and managing the information that goes into them. When they’re not making bots that is. Dear reader, that is only half of the story. Information is managed for a purpose and, in this case, that purpose is providing a service to assist library researchers to answer questions from Members. Some of those questions occasionally finding their way toward team:Anya whenever a query or two is required. And a great many of those questions being Member and Member activity related.

The Members’ Names Information Service frequently gets prodded in service of such questions. But, occasionally, a Member-centric question comes through of a more historical nature and the resources at our disposal aren’t always up to the job. MNIS only really being reliable for Members back to the mid-1980s. We do have our lovely Rush database to call upon but that is more focussed on the social background of people before they became Members. Rather than their activity in Parliament once elected. Which means we frequently find ourselves calling upon Andrew to take out his Wikidata spanner and help formulate SPARQL queries for us.

This week, Andrew popped into pixels to chat to both our crack team of information managing librarians and the crack team of librarians in the Parliament and Constitution Centre. Over the course of a well-spent hour, Andrew gave a general introduction to the more Parliament focussed parts of Wikidata, some of the queries he’s put together and some of the tools he uses to make sure the data doesn’t get vandalised. We hope everyone emerged with a better understanding of what Wikidata is, how useful it can be and how much time and trouble goes into ensuring the information it stores is accurate and untampered with. Should our reader be interested in querying Wikidata with a bias toward UK MPs, Andrew’s even written a handy guide to getting started. Which is a pretty good starting point if you want to learn how to query Wikidata in general. Librarian Anna certainly thinks so. Thanks Andrew, as ever.

Return to public bill mountain

Recent conversations with Legislation Office Liam and Katya pointed to our House of Commons first reading map being rather too busy. That has now been cleared out and is once more back to bare bones.

The fast tracking of second reading was the first to be decanted. This being something that can apparently happen in one of two ways: by an allocation of time motion if no other provisions are required or by a Business of the House motion if they are. What those other provisions might be, how they may impact on other areas of the maps and whether they might lend themselves to enumeration remains a question for next week. Luckily, Legislation Office Huw has kindly offered to step in and lend us his brains. Thanks Huw.

The second thing decanted was examination for hybridity which is not only in its own map but also incorporates feedback from Legislation Office Katya and tidies up a few of our sillier mistakes. Hopefully, next week should see our homework marked and gold stars all round. Hopefully.

Step depth preparations

In procedure editor news, our Jianhan has been preparing the ground for adding step depth to our procedure model. This has been a long, long, long time coming and will - we hope - finally fix the problem with step ordering on both the statutory instrument and treaty websites. A problem we have explained at some considerable length - weeknotes passim. Please forgive a brief recap. Our crack team of librarians capture the date on which a business item took place, but not the time. Partly because capturing the time would be a little more work but mainly because we often lack access to that information or it just isn’t captured. This has led to business items happening on the same day being listed in a fashion that could not be described as ‘procedural’. We’ve seen snaggles such as ‘approved by both Houses’ appearing just above ‘approved by the House of Lords’. So our problems are not limited to the procedurally inappropriate but also the frankly illogical. Jianhan’s work sets the scene for steps to be given a depth in a procedure which will - we hope - be used to order them for display. We’re perfectly well aware that a modern ‘digital service’ sets what it likes to call a ‘roadmap’ and plasters its laptops in ‘Users First’ stickers but we’re pretty happy with a rough plan to make JO Jane happy. A pat on the head being the best we can hope for.

Fettling Rush

Over in the world of the Rush database, Librarian Anna has taken a break from normalising things and backfilled a by-election or two. Which means we now have records for Sarah Green, Jill Mortimer, Anum Qaisar and Kim Leadbeater. Top work, Anna.

Meanwhile, mid-level boss brarian Anya noticed that we had two Members blessed with an honorific of ‘The’ which she found questionable. Upon further investigation, it turned out said Members were also listed with a surname of “O’Conor, O’Conor Don”. Luckily, Wikipedia came to the rescue and we now know that ‘The O’Conor Don’ is a position bestowed on the head of the O’Conor clan. Which may well be Gaelic nobility but - we believe - does not fall into the Rush bucket marked ‘aristocratic’. So the aristocratic checkbox remains unticked. The honorific ‘The’ has been binned and the Don added as a title in the alternative name table. Which we think does the job. Whether or not there might be an O’Somerville Don remains a matter for ancestry.com.

Interesting noticings from around the web

Two hacks this week, neither by us. But both pertinent to parliamentary data nerds everywhere.

Edward has been hard at work combining and comparing directorship entries from the Register of Members’ Financial Interests with declarations to Companies House. All made possible by identifier alignment of MNIS IDs and Companies House officer IDs in Wikidata. Very cool stuff, as we’re sure you’ll agree.

Inspired by Edward’s hacking, Stuart has also been exploring the world of RMFI, this time generating financial league tables. Also well worth a look. Top work both.