Abstract
Over one million tokens of comments were collected for the study using data mining methods. The videos under which the comments were dug out were not chosen arbitrarily but according to the current official national YouTube trends in Poland and the Czechia. The comments were collected under the most popular videos in ten categories: cars, comedy, fashion & lifestyle, gaming, music, non-political interview, politics, report, sports, and video blog. The data collection was carried in 2021–2022 and 2023 from under 40 videos per language and period, 160 videos in total. The corpus data should reveal more about the internet register through the part of speech (POS) frequencies, and the syntactical statistics. In addition, the comments are stylistically clustered in R to make dependencies in linguistic usage visible and better understandable. The study aims to expand the term register, using Polish and Czech, to include language on the internet and distinguish between authentic and non-authentic internet registers in comparison to other registers. An additional sociolinguistic aspect of the analysis is the influence of the Russian war against Ukraine on the linguistic behavior of YouTube users.
Tikhonov, Aleksej