Closed palrogg closed 8 years ago
I like this graph but still too many words per category I think. Do you think it would be possible to reduce each politician to one favorite word, preferably a noun? Or maybe two, a noun and a verb?
It's a good idea to experiment with the clustering. Maybe you want to look at subgroups like the parties in order to get some interesting results.
Closing pitch since story has been opened at #272
I'd like to cluster the speeches held at the Geneva parliament and see what kind of speeches will arise. Or: to perform a TF-IDF search by politician, then plot the results in small multiples.
Headlines
Data source
http://ge.ch/grandconseil/memorial/seances-last/010305
Inspiration
Clustering debates from UK politicians [Blog]
Also interesting:
The Words They Used [NYT]
Words in Politics: Some extensions of the word cloud [Blog]
Choice words [NYT]
Pitch issue checklist
[Pitch]
in the title[Data request]
in the title if I need additional data