A(nother) reminder that our 4th Parliament, data and democracy meetup is on Thursday 14th September at Newspeak House.
There will be interesting talks, nice people and free beer. You should probably come.
Robert and Dia had more meetings preparing for the launch of our new search service.
Samu came up with a clever (if slightly dirty) workaround to allow us to point existing search forms in the general direction of the new search results, hopefully snatching victory from the ever present jaws of defeat.
Wednesday was data day, our fortnightly gabbering session. Lots of people on holiday so all a little quiet. And nothing terribly interesting to report.
And Thursday was a school trip to the Web Science Institute at the University of Southampton.
Much was discussed including:
Architecture and triple store structures
Tools for validating ontologies and instance data
Pros, cons and mitigations of providing public SPARQL endpoints
Google analytics, targeted political advertising, privacy and informed consent
How we might improve the Commons Library taxonomy of subject indexing terms by linking our terms to Wikidata and comparing their class hierarchies to ours
Documenting ontologies in a way that’s more useful to external consumers
Combining search over unstructured documents with search across structured data
How we plan to make our subject headings interoperable with the real-world things they refer to using something a bit like foaf:focus
Aidan met separately with Susan Halford and Mark Weal to plan upcoming research investigating the “Everyday Practices of Semantic Linked Data” across Parliament.
All in all a good day. They did not seem to think we were idiots which, as a great man once said, is not nothing.
Dan and Julie went to the National Archives to share experiences with Alec and Gwyn about change, setting up teams, ways of working, recruitment, development, and more. Their offices are really nice and did not have bed bugs. We were jealous.
Dan started his mission to attend the website team’s ceremonies for the purpose of better communication.
In the joyous world of “long term financial planning” (no idea) Aidan started work on business cases that will cover Data Management, Data Infrastructure, Search and Data Toolkit, and Vocabulary Management.
Aidan also met with Cybersecurity to finalise plans for penetration testing of the new data service.
Jianhan wrote an evidence document and submitted it to the Lords Select Committee on Artificial Intelligence in response to their call for evidence.
Silver and Michael spent all day Tuesday turning select committee sketches into Turtle. Anya popped in between meetings. The sprawling whiteboard sketches are now split into four more fag packet sized models: one for people’s affiliations with committees (members, chairs etc.), one for the core business of select committees (inquiries and remit), one for public submissions of evidence to committees, and a more general purpose one for events like visits, meetings, evidence sessions etc.. They’re still missing useful comments and probably don’t make a lot of sense. This is a job for Anya and Michael next week.
At the end of the day Colin came over to chat about the website team’s thinking on surfacing votes (divisions) and oral contributions on member pages. Anya took Colin, Silver and Michael through the data structure we get from Hansard and the various kinds of contributions and interventions we see there.
Liz wrote up some analysis for the user researchers based on a website survey of the information people were looking for on members during dissolution and parliament.
Matias has been investigating how the Pass Office feed determines an “insert”, “update” or “delete”, in the hope of understanding how we can enhance the process.
He’s also developed a new version of the Event Management System integration which moves invoices from events management into payment systems. The new version has more explicit error handling, allowing it to generate more useful error messages.
David has been helping with extracting data from the building location database (what rooms there are and who’s in them) before the system is retired.
Noel has been analysing data on people records to determine if there are patterns to the issues we’ve spotted.
There were no strolls this week :-(