English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Language models are the new state-of-the-art natural language processing(NLP) models and they are being increasingly used in many NLP tasks. Eventhough there is evidence that language models are biased, the impact of thatbias on the fairness of downstream NLP tasks is still understudied.Furthermore, despite that numerous debiasing methods have been proposed in theliterature, the impact of bias removal methods on the fairness of NLP tasks isalso understudied. In this work, we investigate three different sources of biasin NLP models, i.e. representation bias, selection bias and overamplificationbias, and examine how they impact the fairness of the downstream task oftoxicity detection. Moreover, we investigate the impact of removing thesebiases using different bias removal techniques on the fairness of toxicitydetection. Results show strong evidence that downstream sources of bias,especially overamplification bias, are the most impactful types of bias on thefairness of the task of toxicity detection. We also found strong evidence thatremoving overamplification bias by fine-tuning the language models on a datasetwith balanced contextual representations and ratios of positive examplesbetween different identity groups can improve the fairness of the task oftoxicity detection. Finally, we build on our findings and introduce a list ofguidelines to ensure the fairness of the task of toxicity detection.

On Bias and Fairness in NLP: Investigating the Impact of Bias and Debiasing in Language Models on the Fairness of Toxicity Detection