doi of v1.0.0: 10.5281/zenodo.10901111

This repository contains the code and materials for my Master's Thesis project named Towards Gender-Inclusivity in Danish NLP: Quantifying Bias in Toxicity Detection and Sentiment Analysis, which is going to be available at the library, at which point I will link to it.

• The folder sentiment contains the data and code for the analysis of gender bias in three Danish off-the-shelf sentiment analysis models.
• The folder toxicity_detection contains the data and code for implementing and analyzing Danish toxicity classifiers, and two data augmentation mitigation techniques.

Note that Q or queer is used to describe the group of non-cisgender genders that are referred to in the thesis as the non-conform gender identities.