Abstract

Online misogyny, a category of online abusive language, has serious and harmful social consequences. Automatic detection of misogynistic language online, while imperative, poses complicated challenges to both data gathering, data annotation, and bias mitigation, as this type of data is linguistically complex and diverse. We design an annotation task on labeling Danish social media post with the presence of a comprehensive taxonomy, codebook and weekly discussion rounds.

We reflect on challenges, frontiers and future directions of the annotation task. Further, I will give a general insight in automatic detection of misogyny and other forms of online harm.

By PhD Philine Zeinere, IT-University of Copenhagen.