'Kermit Sueyslide': Scraping Orthographic Innovations Bypassing Twitter's Kill Censorship

This research aimed to document linguistic innovations strategically bypassing content moderation censorship based on a typology by Kim et al. (2021). It focused on the word kill and common synonyms by generating a small corpus that identified the frequency of items across selected regions, as well as the conversational context yielding them. Orthographic optical strategies were the most common across the entire corpus, appearing even in scrapes of items predominantly falling under other categories. An issue, however, is that it may prove difficult to study an environment akin to X (now formerly Twitter) because the site has since made their REST API inaccessible. This paper ideates a myriad of ideas to take this project further in response to the change.