Understanding and quantifying the bias introduced by human annotation of data is a crucial problem for trustworthy supervised learning. Recently, a perspectivist trend has emerged in the NLP community, focusing on the inadequacy of previous aggregation schemes, which suppose the existence of a single ground truth. This assumption is particularly problematic for sensitive tasks involving subjective human judgments, such as toxicity detection. To address these issues, we propose a preliminary approach for bias discovery within human raters by exploring individual ratings for specific sensitive topics annotated in the texts. Our analysis’s object focuses on the Jigsaw dataset, a collection of comments aiming at challenging online toxicity identification.

Bias Discovery within Human Raters: A Case Study of the Jigsaw Dataset

Marta Marchiori Manerba;Riccardo Guidotti;Lucia Passaro;Salvatore Ruggieri
2022-01-01

Abstract

Understanding and quantifying the bias introduced by human annotation of data is a crucial problem for trustworthy supervised learning. Recently, a perspectivist trend has emerged in the NLP community, focusing on the inadequacy of previous aggregation schemes, which suppose the existence of a single ground truth. This assumption is particularly problematic for sensitive tasks involving subjective human judgments, such as toxicity detection. To address these issues, we propose a preliminary approach for bias discovery within human raters by exploring individual ratings for specific sensitive topics annotated in the texts. Our analysis’s object focuses on the Jigsaw dataset, a collection of comments aiming at challenging online toxicity identification.
2022
979-10-95546-98-6
File in questo prodotto:
File Dimensione Formato  
2022.nlperspectives-1.4.pdf

accesso aperto

Tipologia: Versione finale editoriale
Licenza: Creative commons
Dimensione 1 MB
Formato Adobe PDF
1 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11568/1154540
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact