This repository has the implementation of a sexism detection research [1], using a dataset built using the https://sexismo.vercel.app site.
The dataset can be found at /data/labeled-comments.csv.
The file sexism-detection.ipynb contains all the feature selection, training and testing implementation.
The file baseline.ipynb has the implementation of a hate speech detection work made by Davidson et al. [2] and the file bert.ipynb implements the BERT model, to compare results with the current work.
1 - Braga, M. L. P., Nakamura, F. G., and Nakamura, E. F. (2020). Criação e caracterização de um corpus de discurso sexista em português. In Anais do IX Brazilian Workshop on Social Network Analysis and Mining, pages 97–107. SBC. 2 - Davidson, T., Warmsley, D., and Macy, M. (2017). Automated hate speech detection andthe problem of offensive language. Eleventh International AAAI Conference on Weband Social Media.