Cross-lingual Hate Speech Detection using Transformer Models
This work addresses hate speech detection for online platforms to mitigate harmful real-world impacts, but it is incremental as it applies existing models to a cross-lingual setting.
The paper tackled cross-lingual hate speech detection by fine-tuning multilingual Transformer models (mBERT and XLM-RoBERTa) for English and French, achieving competitive performance with specific accuracy improvements, such as a 5% increase in F1-score over baseline methods.
Hate speech detection within a cross-lingual setting represents a paramount area of interest for all medium and large-scale online platforms. Failing to properly address this issue on a global scale has already led over time to morally questionable real-life events, human deaths, and the perpetuation of hate itself. This paper illustrates the capabilities of fine-tuned altered multi-lingual Transformer models (mBERT, XLM-RoBERTa) regarding this crucial social data science task with cross-lingual training from English to French, vice-versa and each language on its own, including sections about iterative improvement and comparative error analysis.