ANDES at SemEval-2020 Task 12: A jointly-trained BERT multilingual model for offensive language detection
This addresses the problem of detecting offensive language across multiple languages for social media and content moderation, but it is incremental as it builds on existing BERT methods.
The paper tackled multilingual offensive language detection by jointly fine-tuning a single Multilingual BERT model across five languages, achieving competitive results close to top-performing systems with shared parameters.
This paper describes our participation in SemEval-2020 Task 12: Multilingual Offensive Language Detection. We jointly-trained a single model by fine-tuning Multilingual BERT to tackle the task across all the proposed languages: English, Danish, Turkish, Greek and Arabic. Our single model had competitive results, with a performance close to top-performing systems in spite of sharing the same parameters across all languages. Zero-shot and few-shot experiments were also conducted to analyze the transference performance among these languages. We make our code public for further research