Abstract

Disparate biases associated with datasets and trained classifiers in hateful and abusive content identification tasks have raised many concerns recently. Although the problem of biased datasets on abusive language detection has been addressed more frequently, biases arising from trained classifiers have not yet been a matter of concern. In this paper, we first introduce a transfer learning approach for hate speech detection based on an existing pre-trained language model called BERT (Bidirectional Encoder Representations from Transformers) and evaluate the proposed model on two publicly available datasets that have been annotated for racism, sexism, hate or offensive content on Twitter. Next, we introduce a bias alleviation mechanism to mitigate the effect of bias in training set during the fine-tuning of our pre-trained BERT-based model for hate speech detection. Toward that end, we use an existing regularization method to reweight input samples, thereby decreasing the effects of high correlated training set’ s n-grams with class labels, and then fine-tune our pre-trained BERT-based model with the new re-weighted samples. To evaluate our bias alleviation mechanism, we employed a cross-domain approach in which we use the trained classifiers on the aforementioned datasets to predict the labels of two new datasets from Twitter, AAE-aligned and White-aligned groups, which indicate tweets written in African-American English (AAE) and Standard American English (SAE), respectively. The results show the existence of systematic racial bias in trained classifiers, as they tend to assign tweets written in AAE from AAE-aligned group to negative classes such as racism, sexism, hate, and offensive more often than tweets written in SAE from White-aligned group. However, the racial bias in our classifiers reduces significantly after our bias alleviation mechanism is incorporated. This work could institute the first step towards debiasing hate speech and abusive language detection systems.

Highlights

  • DisclaimerThis article uses words or language that is considered profane, vulgar, or offensive by some readers

  • Hate speech detection and racial bias mitigation in social media based on BERT model justified but neither we nor PLOS in any way endorse the use of these words or the content of the quotes

  • Considering the disparate distribution of tweets in different classes described in Table 1, it is justifiable that we are dealing with imbalanced datasets

Read more

Summary

Introduction

DisclaimerThis article uses words or language that is considered profane, vulgar, or offensive by some readers. Hate speech detection and racial bias mitigation in social media based on BERT model justified but neither we nor PLOS in any way endorse the use of these words or the content of the quotes. Owning to the recent proliferation of user-generated textual contents in online social media, a wide variety of studies have been dedicated to investigating these contents in terms of hate or toxic speech, abusive or offensive languages, etc., [1,2,3,4,5,6]. With regard to the mobility and anonymous environment of online social media, suspect users, who generate abusive contents or organize the hate-based activities, exploit these online platforms to propagate hate and offensive contents towards other users and communities [2, 7]; where it leads to personal trauma, hate crime, cyber-bullying, and discrimination (mainly racial and sexual discriminations) [8]. Online social media have been persuaded to define policies to remove such harmful content from their platforms since 2015 [9, 10]

Methods
Results
Conclusion

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.