Pre-training language models for identifying patronizing and condescending language: an analysis

Perez Almendros, Carla ORCID: https://orcid.org/0000-0001-9360-4011, Espinosa-Anke, Luis ORCID: https://orcid.org/0000-0001-6830-9176 and Schockaert, Steven ORCID: https://orcid.org/0000-0002-9256-2881 2022. Pre-training language models for identifying patronizing and condescending language: an analysis. Presented at: LREC 2022, Marseille; France, 20-25 June 2022. Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022). European Language Resources Association, pp. 3902-3911.

PDF - Published Version Available under License Creative Commons Attribution Non-commercial. Download (207kB)

Abstract

Patronizing and Condescending Language (PCL) is a subtle but harmful type of discourse, yet the task of recognizing PCL remains under-studied by the NLP community. Recognizing PCL is challenging because of its subtle nature, because available datasets are limited in size, and because this task often relies on some form of commonsense knowledge. In this paper, we study to what extent PCL detection models can be improved by pre-training them on other, more established NLP tasks. We find that performance gains are indeed possible in this way, in particular when pre-training on tasks focusing on sentiment, harmful language and commonsense morality. In contrast, for tasks focusing on political speech and social justice, no or only very small improvements were witnessed. These findings improve our understanding of the nature of PCL.

Item Type:	Conference or Workshop Item (Paper)
Status:	Published
Schools:	Professional Services > Advanced Research Computing @ Cardiff (ARCCA) Schools > Computer Science & Informatics
Publisher:	European Language Resources Association
ISBN:	979-109554672-6
Date of First Compliant Deposit:	25 May 2022
Date of Acceptance:	4 April 2022
Last Modified:	27 Feb 2025 14:42
URI:	https://orca.cardiff.ac.uk/id/eprint/150046

Actions (repository staff only)

Edit Item