-
Full report here as a Google Doc for some reason: docs.google.com/document/d/1HKK_ordzVVuzBjVBSmPnPaSyHc-oqpC0aBgSkxYUrjE/edit - This uses a machine-learning program to decide what words count as 'toxic' without context. - The program thinks calling someone 'Tory' is toxic. - Only 5% of tweets analysed were 'toxic'. t.co/r6EYUk7AkAOn Internet Archive
♻️ 2 Retweets ❤️ 11 Favorites
-
'corrupt' is also a prominent word identified with toxicity. Calling a member of the current Government 'corrupt' is a statement of fact.Permalink On Internet Archive
-
The Tories are responsible for 190,000 unnecessary deaths (not even including those from austerity) but god forbid someone on a website calls them *checks report* "fools".Permalink On Internet Archive
-
Maybe using a computer program that can't understand context or even the semantic meaning of a sentence is not the best way to conduct this kind of research. @Robert_Gordon_/1590303137793212417Permalink On Internet Archive