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

Abstract

With the epidemic continuing, hatred against Asians is intensifying in countries outside Asia, especially among the Chinese. Thus, 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 that is an extension of the anti-Asian hate speech dataset on Twitter, including 2,035 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 nonhate tweets in the dataset. Second, we finetune the BERT models based on the relevant datasets, and demonstrate strategies including 1) cleaning the hashtags, usernames being @, URLs, and emojis before the fine-tuning process, and 2) training with the data while validating with the "clean" data (and the opposite) are not effective for improving performance. Third, we investigate the performance of advanced fine-tuning strategies with 1) modelcentric approaches, such as discriminative finetuning, gradual unfreezing, and warmup steps, and 2) data-centric approaches, which incorporate data trimming and data augmenting, and show that both strategies generally improve the performance, while data-centric ones outperform the others, which demonstrate the feasibility and effectiveness of the data-centric approaches. 2 The hate crime data for 2020 is available on the FBI's Crime Data Explorer: https: //crime-data-explorer.app.cloud.gov 3 See https://www.who.int/docs/ default-source/coronaviruse/ covid19-stigma-guide.pdf