Originally published at: https://boingboing.net/2019/03/02/julia-reda-washington-dc.html
…
I wonder whether it would be possible to use spelling to make an educated guess as to out what percentage of the English language tweets were coming from the US? As opposed to the UK or other commonwealth countries.
Given the limited text to work with in tweets, it would have to be something fairly crude, e.g. if the spelling was correct, it didn’t come from the United States. (Which I’m actually pretty certain would, in fact, be a reliable indicator, I’m sorry to say.)
Well certainly if spellings such as “colour” are used, it implies the writer is from outside the US, although due to spell- checkers aggressively promoting US- normal spellings, the opposite isn’t always true. Slang could also be used, but I suspect the authors of this study would find proper accuracy to be… not useful to their goals, shall we say.
It could also be that people are more comfortable writing in English than their native language due to many years spent abroad and other issues I’d rather not dwell on.
They seem to have removed it.
This topic was automatically closed after 5 days. New replies are no longer allowed.