An unbiased offensive text detection method based on BERT and sentiment analysis.

Saved in:

Bibliographic Details
Title:	An unbiased offensive text detection method based on BERT and sentiment analysis.
Authors:	YUAN, Liang¹, GUO, Weibi¹ gweibin@ecust.edu.cn
Source:	Computer Engineering & Science / Jisuanji Gongcheng yu Kexue. May2026, Vol. 48 Issue 5, p906-913. 8p.
Subjects:	Sentiment analysis, Data augmentation, Long short-term memory, Discrimination (Sociology), Language models, Natural language processing
Abstract:	Offensive information on the internet poses severe harm to individuals and society. In offensive text detection methods, existing methods suffer from misjudging non-offensive texts containing profanity and bias against special groups. To address the former issue, this paper proposes a sentiment analysis-based offensive text detection (SAOD) model, which uses sentiment features to assist in predicting whether a text is offensive. To tackle the latter issue, a debiasing data augmentation method called special groups mask (SGM) is proposed. This method masks special groups during training, ensuring that special groups are not directly involved in model training, thereby reducing the model's bias towards these groups. Using BERT+LSTM as the base model, experiments were conducted on publicly available datasets ToxiCN and COLD. The experimental results show that the former method improved the base model's F1-score from 80.18% to 82.67%. Based on this, the latter method reduces the false positive rate (FPR) from 18.27% to 12.77%. [ABSTRACT FROM AUTHOR]
	Copyright of Computer Engineering & Science / Jisuanji Gongcheng yu Kexue is the property of Computer Engineering & Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	Offensive information on the internet poses severe harm to individuals and society. In offensive text detection methods, existing methods suffer from misjudging non-offensive texts containing profanity and bias against special groups. To address the former issue, this paper proposes a sentiment analysis-based offensive text detection (SAOD) model, which uses sentiment features to assist in predicting whether a text is offensive. To tackle the latter issue, a debiasing data augmentation method called special groups mask (SGM) is proposed. This method masks special groups during training, ensuring that special groups are not directly involved in model training, thereby reducing the model's bias towards these groups. Using BERT+LSTM as the base model, experiments were conducted on publicly available datasets ToxiCN and COLD. The experimental results show that the former method improved the base model's F1-score from 80.18% to 82.67%. Based on this, the latter method reduces the false positive rate (FPR) from 18.27% to 12.77%. [ABSTRACT FROM AUTHOR]
ISSN:	1007130X
DOI:	10.3969/j.issn.1007-130X.2026.05.014