# Speech Detection Task Against Asian Hate: BERT the Central, While Data-Centric Studies the Crucial

arXiv: https://arxiv.org/abs/2206.02114 · HTML: https://mlsanity.com/paper/2206.02114/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 64 references that are in the map.
- First posted on arXiv 5 Jun 2022.

## Builds on (most cited first, 23 of 23)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation](https://mlsanity.com/paper/1609.08144/) (2016)
- [How transferable are features in deep neural networks?](https://mlsanity.com/paper/1411.1792/) (2014)
- [On Layer Normalization in the Transformer Architecture](https://mlsanity.com/paper/2002.04745/) (2020)
- [Evaluation: from precision, recall and F-measure to ROC, informedness, markedness and correlation](https://mlsanity.com/paper/2010.16061/) (2020)
- [How to Fine-Tune BERT for Text Classification?](https://mlsanity.com/paper/1905.05583/) (2019)
- [HateXplain: A Benchmark Dataset for Explainable Hate Speech Detection](https://mlsanity.com/paper/2012.10289/) (2020)
- [FinBERT: Financial Sentiment Analysis with Pre-trained Language Models](https://mlsanity.com/paper/1908.10063/) (2019)
- [Revisiting Few-sample BERT Fine-tuning](https://mlsanity.com/paper/2006.05987/) (2020)
- [HateBERT: Retraining BERT for Abusive Language Detection in English](https://mlsanity.com/paper/2010.12472/) (2020)
- [Multilingual and Multi-Aspect Hate Speech Analysis](https://mlsanity.com/paper/1908.11049/) (2019)
- [What Would Elsa Do? Freezing Layers During Transformer Fine-Tuning](https://mlsanity.com/paper/1911.03090/) (2019)
- [Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach](https://mlsanity.com/paper/2010.07835/) (2020)
- [Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning](https://mlsanity.com/paper/2102.03983/) (2021)
- [Learning from Rules Generalizing Labeled Exemplars](https://mlsanity.com/paper/2004.06025/) (2020)
- [Detecting East Asian Prejudice on Social Media](https://mlsanity.com/paper/2005.03909/) (2020)
- [Racism is a Virus: Anti-Asian Hate and Counterspeech in Social Media during the COVID-19 Crisis](https://mlsanity.com/paper/2005.12423/) (2020)
- [Detection of Hate Speech using BERT and Hate Speech Word Embedding with Deep Model](https://mlsanity.com/paper/2111.01515/) (2021)
- [Transformer-based Language Model Fine-tuning Methods for COVID-19 Fake News Detection](https://mlsanity.com/paper/2101.05509/) (2021)
- [Cyberbullying Classifiers are Sensitive to Model-Agnostic Perturbations](https://mlsanity.com/paper/2201.06384/) (2022)
- [Hate speech detection using static BERT embeddings](https://mlsanity.com/paper/2106.15537/) (2021)
- [A Feature Extraction based Model for Hate Speech Identification](https://mlsanity.com/paper/2201.04227/) (2022)

## Abstract (arXiv)

With the COVID-19 pandemic continuing, hatred against Asians is intensifying in countries outside Asia, especially among the Chinese. There is an urgent need to detect and prevent hate speech towards Asians effectively. In this work, we first create COVID-HATE-2022, an annotated dataset including 2,025 annotated tweets fetched in early February 2022, which are labeled based on specific criteria, and we present the comprehensive collection of scenarios of hate and non-hate tweets in the dataset. Second, we fine-tune the BERT model based on the relevant datasets and demonstrate several strategies related to the "cleaning" of the tweets. Third, we investigate the performance of advanced fine-tuning strategies with various model-centric and data-centric approaches, and we show that both strategies generally improve the performance, while data-centric ones outperform the others, and it demonstrates the feasibility and effectiveness of the data-centric approaches in the associated tasks.