ontologies

2020 - Week 24

This week felt like another of those where lots of things happened but somehow there’s not an awful lot to show. Early efforts at mapping super affirmative procedure proceeded no further, but at least we now have a clarifying email from Journal Office Jane. All the calculations for the egg timer were polished off, but none of them seemed to work. This is life with parliamentary procedure and computers we fear. All we can really do is keep on keeping on.

Procedure mapping

Last week we were all about enthusiasm for finally mapping the super affirmative procedure(s?). But enthusiasm is no substitute for understanding. So that work took a pause this week, and Librarian Jayne and self-titled computational expert Michael decided they’d take a look at the information architecture of Jayne’s SPARQL query pages instead. Information architecture being what pompous computational blokes call filing. And librarians call librarianship. Pages were reorganised, queries re-filed and we think the thing makes more sense. We now have a page that links to instrument types, be they Proposed Negative Statutory Instruments, Statutory Instruments or Treaties. The pages are not yet linked to from the query homepage - it’s not quite but almost a website. A meeting was scheduled for Friday afternoon to sort out this oversight, but unfortunately Jayne’s plumbing exploded. Not wishing to alarm anyone, we should note that the plumbing is of the house she lives in. Personal plumbing explosions being a problem that remains firmly in Michael’s camp. Next week offers the attractions of yet more filing and more super affirmative procedure. And if that is not a good reason to get out of bed, we don’t know what is.

In other procedural news, we’ve had a long-standing problem with House of Lords business disappearing over the event horizon. Today’s business, and business on a future day, can be accessed at a URL and each event can be navigated to - for example a page for a motion. Which librarians can link to. For some reason, once the day is done, the route to the event page disappears from the web never to be seen again. Only a long and lonely PDF remains. This is a problem for our librarian contingent. If, for any reason, they fall a day behind something happening in the Lords, it is missed forever and a link to the big bag of stuff that is a PDF must be made instead. Which makes us sad. So Librarian Anya asked her computational colleagues to deploy their advanced screen scraping skills to grab a copy of Lords’ business at dawn and cache it somewhere before it disappears. Ever willing to please, Robert has taken on this task and is now pointing screen scraping code at the source of the problem.

All about the collaboration

On Friday, Anya and Michael met with Nick from the Participation Team. Nick works on Committee outreach, organising visits across the UK to bring underrepresented communities together with committee members. To aid him in his work, he’s started to build a rudimentary customer relationship management system in the form of an Excel spreadsheet. Nick’s spreadsheet captures organisations, their geographic operating areas, the sectors they operate in and the specialisms they cover. He’d heard about the work of the Indexing and Data Management Section from a colleague, and wanted some tips and techniques on how best to catalogue and categorise. Anya promised to send him some of the Library indexing principles, and to help him come up with a suitable vocabulary for categorising. Michael committed to taking a look at the spreadsheet in an attempt to make it more relational and better manage “reference data”. For want of a better word. Michael also contacted Gordon to check how committee teams handle CRM for evidence givers. In assorted bits of “Word, Excel [and] OneNote” came the not completely reassuring answer.

Way back at the start of May, our friends in Oxford got in touch to say they’d updated their ParlRules Data on House of Commons public Standing Orders. The new data release has additional text cleaning and is “now organised around whole Standing Orders, rather than the more disaggregated level of sub-articles.” Having been distracted by egg timers Michael re-focussed, pulled out his finger and put the data through our normal normalisation script and pushed it to our Standing Order prototype application. He’s still a little confused about what the move from sub-articles to whole orders means in practice, and failed to detect any discernible difference in the shape of the data. But, we have a call to Oxford pencilled in for Monday. So, hopefully, all will become clear. Or clearer.

Your weekly egg timer update

If brow wrinkles were proof of work, Robert and Michael would be bitcoin billionaires by now. This week saw their foreheads age by a decent decade as they once more poured over egg timer calculations and, with only minimal swearing, expanded out our monkey patched date class. The good news is the controller is now thinner and the model fatter; this is a thing our grey beard friends tend to applaud. And who are we to doubt elders and betters. There is now calculation code for all nine of our supported variations of procedure. So that’s progress.

In more good news, the egg timer now returns all the right dates. Just not necessarily for the right instruments. That’s a problem. Next week promises a long, slow trek through the dank valleys of debugging. There are quite a lot of moving parts here - from calendar management to calendar synchronisation to the calculations themselves - so we assume Robert and Michael’s wee faces will be more wrinkle than flesh by the end of the week.

Tiny chisel news

Thursday morning saw a pixel based meeting between Anya, Lorna, Cassie, Oli, Phillip, Robert and Michael to chat though Economic Indicators and similar library services. It was the first time Anya, Michael and Robert had done any proper domain modelling since lockdown began. It’s definitely easier and more of a pleasure with a whiteboard and warm human bodies in the room, but Michael’s attempts at Omnigraffle over a shared screen went better than expected. The picture produced needs to be compared to existing research service models and any gaps firmly plugged. So that’s a job for next week perhaps.

Tidying the record model took up a not inconsiderable portion of Anya, Robert and Michael’s week. A new version is now published but it still needs work. Particularly around revision management, researcher specialisms and capturing publication cycles. All useful things that dropped out of the meeting with Lorna and colleagues.