trumpogram
The World, according to Donald Trump. Each country's size is approximately proportional to the number of times that it was mentioned in Trump's tweets and campaign speeches.
Final Output
About
Inspired by this tweet: https://twitter.com/SwiftOnSecurity/status/832074664193249281
There isn't much to this project, just plugging a few existing tools together.
The identification of place names is performed using the spacy python library, so any issues in this area aren't my problem :P.
I used spacy to output a list of GPEs and NORPs (Geo-Political Entities, Nationalities Or Religious/Political Groups).
The script then resolves "synonyms", and generates a column of quantities, corresponding to the
countries in the geographical data.
Note: Mentions of individual US states are not included in the US total. Otherwise it would probably be way too big...
Corpus
The corpus for this project was obtained from multiple sources (I might add more):
- 2016 Campaign speeches, from ryanmcdermott/trump-speeches
- @realDonaldTrump's tweets, via tweet_dumper.py
Mapping
I obtained the map data from thematicmapping.org I generated the final map using the ScapeToad tool.