HULAT at SemEval-2023 Task 10: Data augmentation for pre-trained transformers applied to the detection of sexism in social media

TL;DR

This paper investigates the effectiveness of data augmentation techniques combined with transformer models like RoBERTa for detecting sexism in social media, achieving best results on certain subtasks but with room for improvement.

Contribution

It explores the impact of data augmentation on transformer-based models for sexism detection, providing insights into their performance across different subtasks.

Findings

RoBERTa with data augmentation improved results for tasks B and C

Synthetic data did not enhance performance for task C

The approach shows potential but needs further refinement for fine-grained classification

Abstract

This paper describes our participation in SemEval-2023 Task 10, whose goal is the detection of sexism in social media. We explore some of the most popular transformer models such as BERT, DistilBERT, RoBERTa, and XLNet. We also study different data augmentation techniques to increase the training dataset. During the development phase, our best results were obtained by using RoBERTa and data augmentation for tasks B and C. However, the use of synthetic data does not improve the results for task C. We participated in the three subtasks. Our approach still has much room for improvement, especially in the two fine-grained classifications. All our code is available in the repository https://github.com/isegura/hulat_edos.