MedAI at SemEval-2021 Task 5: Start-to-end Tagging Framework for Toxic Spans Detection
Zhen Wang | Hongjie Fan | Junfei Liu
Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021)
This paper describes the system submitted to SemEval 2021 Task 5: Toxic Spans Detection. The task concerns evaluating systems that detect the spans that make a text toxic when detecting such spans are possible. To address the possibly multi-span detection problem, we develop a start-to-end tagging framework on top of RoBERTa based language model. Besides, we design a custom loss function that takes distance into account. In comparison to other participating teams, our system has achieved 69.03% F1 score, which is slightly lower (-1.8 and -1.73) than the top 1(70.83%) and top 2 (70.77%), respectively.