UIT-ISE-NLP at SemEval-2021 Task 5: Toxic Spans Detection with BiLSTM-CRF and ToxicBERT Comment Classification

TL;DR

This paper describes a system combining BiLSTM-CRF and ToxicBERT for toxic span detection in online comments, achieving a 62.23% F1-score, advancing automated toxicity identification.

Contribution

The paper introduces a hybrid BiLSTM-CRF and ToxicBERT approach specifically designed for toxic span detection in social media posts.

Findings

Achieved 62.23% F1-score on the Toxic Spans Detection task.

Demonstrated effectiveness of combining sequence labeling with classification models.

Provided a new baseline for toxic span detection in SemEval-2021.

Abstract

We present our works on SemEval-2021 Task 5 about Toxic Spans Detection. This task aims to build a model for identifying toxic words in whole posts. We use the BiLSTM-CRF model combining with ToxicBERT Classification to train the detection model for identifying toxic words in posts. Our model achieves 62.23% by F1-score on the Toxic Spans Detection task.