TækTåk: Syntactic Analysis of Language Use on Danish TikTok

Abstract

Language use is different across different language communities. Social media provides a rich source for studying how language varies, as it contains large data for a wide variety of sub-communities. In this paper, we study language usage on Danish TikTok. TikTok is a video-based platform, but most users are mainly active in the text-based comment sections. With the goal of analyzing language usage on this language variety, we contribute: 1) the first Danish social media treebank annotated for Universal Dependencies 2) evaluation of a variety of parsers using the new treebank, showing that cross-lingual in-domain data provides a valuable signal 3) a comparison of syntactic trends on standard Danish languages and TikTok language.

Resources

Details