I've done various 'pet projects' over the years in this realm, from exploring political bias in word embeddings to simple voting maps showing regional shifts, to comparisons of actual crime vs fear of crime across various demographics. I've even gotten to partner up with actual journalism outlets.
I love doing this stuff but I'd like to collaborate with others (either with technical or promotional capacity) to make these kinds of tools more useful, relevant, and ultimately actually inform people better.
Would love to have an open-database on political representatives. Some all-in-one source of info:
- Where do they get their funding?
- Their assets
- Their political stances: voting records, statements to the press (multimedia: video, audio, press excerpts)
- Their relationships: who their wife or children are working for? Who are they friend with (seems intrusive but it has a lot of influence on how they vote)
The UI would be important, something easily accessible, searchable, dynamic.
This is hard work and with no hope of monetizing it so I guess its a lost cause but that would allow to have better insights when casting your own vote.
I think this is more relevant in the US... in party-based systems it becomes much more interesting to see what the party trends are... it's kind of the same types of information but it's even harder to get at because it's spread out over a party. But it's also a little easier to see because parties have to coordinate internally and artifacts of that coordination are pretty easily available through public channels. But yeah, no hope of monetizing so a hobby it will stay.
(https://nimishg.com/showcase/) though I recently migrated the site so I'm actually not totally sure all the showcase links work. All of the code is on github (https://github.com/nimishgautam) but might take a little effort to get working on your own
I've done various 'pet projects' over the years in this realm, from exploring political bias in word embeddings to simple voting maps showing regional shifts, to comparisons of actual crime vs fear of crime across various demographics. I've even gotten to partner up with actual journalism outlets.
I love doing this stuff but I'd like to collaborate with others (either with technical or promotional capacity) to make these kinds of tools more useful, relevant, and ultimately actually inform people better.