It does a good job of pulling out the key entities though.
#dataisthebestdisinfectant
Not the kind of doctor you might want at 30,000 feet. Or the kind of chemist you'd trust to give you the right tablets. Old, Curmudgeonly, Big, Ugly and Welsh. And somewhat dyspraxic.
On the side of the child, not the gun. No matter whose child, or whose gun.
I run the #chem4word project in my spare time.
AKA 'Scope'
| Background | Originally a chemist, but drifted into informatics and then information technology. Now work for a university. |
| Interests | Being snobbish about the Welsh coastline, scientific informatics, science, nature, walking, food, classical music, film, good books, gardens and my adorable choccie Labrador. |
| Projects | https://chem4word.co.uk |
Been messing around some more with MPs' interests. https://www.theyworkforyou.com/interests/ now has downloadable JSON files. I've run these for 2024/5 through a KNIME workflow and done some basic NLP processing to pull out key players.
This, for instance, shows who the Carlton Club happens to support. Nodes are scaled in proportion to the money given/received.
Green links are donations and loans, yellow ones gifts,, benefits and hospitality
So, i've been messing about more with the MP's Interests Visualiser (the 'sunlight engine'). I've noticed that the API returns incomplete data: many interests are missing. I imagine this is having to grapple with legacy data.
However, I've just run the latest version on the Tory Party. Node size is proportional to total donated or received. Edge thickness indicated the size of donations to that individual.
Mr. Jenrick does quite well out of it all.
There's other kinds of fun to be had. This is what our dear Prime Minister has been given in the way of tickets recently:
I'm going to do some more investigations and post them here when I find something new and interesting.
If you want to know more about the Engine then please contact me directly via Private Mention