Bias against Black immigrants and Black people among chinese online: an empirical study on the video platform Bilibili and its floating comments

While racism against people of African descent has been well-researched under a western context, the situation of these immigrants in China has been generally overlooked by the academia. Through data-scraping and BERT-based two-step binary classification of Bilibili video floating comments, this research has investigated anti-Black racism on the Chinese internet. This research analyzed the correlation between certain video metadata and negative comments present in the video. The author also manually classified 2000 randomly selected comments based on the type of sentiment expressed, which is used as the training set of the proposed natural language processing model. By collecting and analyzing 90259 floating comments from 683 videos, the study has found that 9.58% of comments were race-related and among these, 80.37% expressed negative sentiments. The author identifies online racism manifesting in the forms of nationalist repulsion of immigration, call for genetic and cultural purity, and associating Black people with negative qualities like violent, womanizing, lazy, and unintelligent. In summary, this research completes a large-scale quantitative and qualitative analysis of the unique types of anti-Black online racism among Chinese netizens while also accounting for the prevalence and causes of such racism, especially in the socio-cultural context of China.