A Simple Voting Mechanism for Online Sexist Content Identification
This work addresses the problem of identifying online sexist content for social media moderation, but it is incremental as it builds on existing language models with a simple voting approach.
The paper tackled sexism detection in social media by combining BERT models with a voting mechanism, achieving the best results among their models, though specific performance numbers are not provided.
This paper presents the participation of the MiniTrue team in the EXIST 2021 Challenge on the sexism detection in social media task for English and Spanish. Our approach combines the language models with a simple voting mechanism for the sexist label prediction. For this, three BERT based models and a voting function are used. Experimental results show that our final model with the voting function has achieved the best results among our four models, which means that our voting mechanism brings an extra benefit to our system. Nevertheless, we also observe that our system is robust to data sources and languages.