>>22221274What a waste of time that was. They admitted that their information aggregation tool couldn't parse /pol/'s jargon and that they thus excluded hundreds of thousands of their posts from the dataset, then, without apparent reflection, concluded that /pol/ largely discusses mainstream topics and has little influence outside of itself. Realizing that this made /pol/ look extremely bland, which would render their already questionable study entirely uninteresting, they then decided to do a comparative analysis of its poorly-defined toxicity and insult quotients so they can conclude that it's full of angry racists. The bit where they dance around why the term "animal" is rated as highly insulting is, unintentionally, very funny. They probably spent tens of thousands of dollars to give an extremly mathematically complex solution to questions anyone could answer by observing the websites they studied for about 10 minutes each. An utter crock of shite.
The most interesting thing in there is that there's apparently been a dedicated 4chan post grabber called 4CAT since 2022 (
https://4cat.nl/). I wonder if these researchers realize that there's home-grown archives they could look at.
