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

Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021), pages 846–851 Bangkok, Thailand (online), August 5–6, 2021. ©2021 Association for Computational Linguistics 846 UIT-ISE-NLP at SemEval-2021 Task 5: Toxic Spans Detection with BiLSTM-CRF and ToxicBERT Comment Classification Son T. Luu University of Information Technology Vietnam National University Ho Chi Minh City, Vietnam sonlt@uit.edu.vn Ngan Nguyen University of Information Technology Vietnam National University Ho Chi Minh City, Vietnam ngannlt@uit.edu.vn 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. 1 Introduction Detecting toxic posts on social network sites is a crucial task for social media moderators in order to keep a clean and friendly space for online dis- cussion. To identify whether a comment or post is toxic or not, social network administrators often read the whole comment or post. However, with a large number of lengthy posts, the administrators need assistance to locate toxic words in each post to decide whether a post is toxic or non-toxic instead of reading the whole post. The SemEval-2021 Task 5 (Pavlopoulos et al., 2021) provides a valuable dataset called Toxic Spans Detection dataset in or- der to train the model for detecting toxic words in lengthy posts. Based on the dataset from the shared task, we implement the machine learning model for detect- ing toxic words posts. Our model includes: the BiLSTM-CRF model (Lample et al., 2016) for de- tecting the toxic spans in the post, and the Toxi- cBERT (Hanu and Unitary team, 2020) for classify- ing whether the post is toxic or not. Before training the model, we pre-process texts in posts and encode them by the GloVe word embedding (Pennington et al., 2014). Our model achieves 62.23% on the test set provided by the task organizers. 2 Related works Many corpora are constructed for toxic speech de- tection problems. They consist of flat label and hierarchical label datasets. The flat label datasets only classify one label for each comment in the dataset (e.g., hate, offensive, clean), while hierar- chical datasets can classify multiple aspects of the comment (e.g., hate about racism, hate about sexual oriented, hate about religion, and hate about dis- ability). For flat label, we present several datasets including the two datasets which are provided by Waseem and Hovy (2016) and Davidson et al. (2017) in English, the dataset of Albadi et al. (2018) in Arabic, and the dataset of by Alfina et al. (2017) in Indonesian. For the hierarchical label, we introduce the dataset constructed by Zampieri et al. (2019) in English, the dataset provided by Fortuna et al. (2019) in Portuguese, and the CO- NAN dataset by Chung et al. (2019), which is the multilingual corpus (constructed in Italian, English, and French). In addition, many of …

Content truncated to 500 words