Syntax and prejudice: ethically-charged biases of a syntax-based hate speech recognizer unveiled

Michele Mastromattei; Leonardo Ranaldi; Francesca Fallucchi; Fabio Massimo Zanzotto

doi:10.7717/peerj-cs.859

Syntax and prejudice: ethically-charged biases of a syntax-based hate speech recognizer unveiled

PeerJ Comput Sci. 2022 Feb 3:8:e859. doi: 10.7717/peerj-cs.859. eCollection 2022.

Authors

Michele Mastromattei¹, Leonardo Ranaldi², Francesca Fallucchi², Fabio Massimo Zanzotto¹

Affiliations

¹ Department of Enterprise Engineering, University of Roma "Tor Vergata", Rome, Italy.
² Department of Innovation and Information Engineering, Guglielmo Marconi University, Rome, Italy.

Abstract

Hate speech recognizers (HSRs) can be the panacea for containing hate in social media or can result in the biggest form of prejudice-based censorship hindering people to express their true selves. In this paper, we hypothesized how massive use of syntax can reduce the prejudice effect in HSRs. To explore this hypothesis, we propose Unintended-bias Visualizer based on Kermit modeling (KERM-HATE): a syntax-based HSR, which is endowed with syntax heat parse trees used as a post-hoc explanation of classifications. KERM-HATE significantly outperforms BERT-based, RoBERTa-based and XLNet-based HSR on standard datasets. Surprisingly this result is not sufficient. In fact, the post-hoc analysis on novel datasets on recent divisive topics shows that even KERM-HATE carries the prejudice distilled from the initial corpus. Therefore, although tests on standard datasets may show higher performance, syntax alone cannot drive the "attention" of HSRs to ethically-unbiased features.

Keywords: Bias; Explainability; Hate speech; Neural networks; Syntax.

Grants and funding

This research was funded by the 2019 BRIC INAIL ID32 SfidaNow project. There was no additional external funding received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.