MarIA and BETO are sexist: evaluating gender bias in large language models for Spanish.

The study of bias in language models is a growing area of work, however, both research and resources are focused on English. In this paper, we make a first approach focusing on gender bias in some freely available Spanish language models trained using popular deep neural networks, like BERT or RoBER...

Full description

Bibliographic Details
Published in:Language Resources & Evaluation Vol. 58; no. 4; pp. 1387 - 1418
Main Authors: Garrido-Muñoz, Ismael, Martínez-Santiago, Fernando, Montejo-Ráez, Arturo
Format: Article
Published: Springer Nature Dec2024
Subjects:
Online Access:View this record in EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=180627302&site=ehost-live
header:
  @attributes:
    shortDbName: hlh
    uiTerm: 180627302
    longDbName: Humanities International Complete
    uiTag: AN
  controlInfo:
    bkinfo:
    jinfo:
      jid:
        1574020X
        179V
      jtl: Language Resources & Evaluation
      issn: 1574020X
      maglogo: N
    pubinfo:
      dt: Dec2024
      vid: 58
      iid: 4
      pid: 237
      pub: Springer Nature
    artinfo:
      ui:
        180627302
        10.1007/s10579-023-09670-3
      ppf: 1387
      ppct: 31
      formats:
        fmt:
          – @attributes:
              type: T
          – @attributes:
              type: P
              size: 1.8MB
      tig:
        atl: MarIA and BETO are sexist: evaluating gender bias in large language models for Spanish.
      aug:
        au:
          Garrido-Muñoz, Ismael
          Martínez-Santiago, Fernando
          Montejo-Ráez, Arturo
        affil: https://ror.org/0122p5f64 CEATIC, Universidad de Jaén, Campus Las Lagunillas, 23071, Jaén, Spain
      su:
        Artificial neural networks
        Language models
        Spanish language
        Deep learning
        Sexism
        Algorithmic bias
      sug:
        subj:
          Artificial neural networks
          Language models
          Spanish language
          Deep learning
          Sexism
          Algorithmic bias
      keyword:
        BERT
        Bias evaluation
        Gender bias
        Language model
        RoBERTa
      ab: The study of bias in language models is a growing area of work, however, both research and resources are focused on English. In this paper, we make a first approach focusing on gender bias in some freely available Spanish language models trained using popular deep neural networks, like BERT or RoBERTa. Some of these models are known for achieving state-of-the-art results on downstream tasks. These promising results have promoted such models' integration in many real-world applications and production environments, which could be detrimental to people affected for those systems. This work proposes an evaluation framework to identify gender bias in masked language models, with explainability in mind to ease the interpretation of the evaluation results. We have evaluated 20 different models for Spanish, including some of the most popular pretrained ones in the research community. Our findings state that varying levels of gender bias are present across these models.This approach compares the adjectives proposed by the model for a set of templates. We classify the given adjectives into understandable categories and compute two new metrics from model predictions, one based on the internal state (probability) and the other one on the external state (rank). Those metrics are used to reveal biased models according to the given categories and quantify the degree of bias of the models under study.
      pubtype: Academic Journal
      doctype: Article
      src: R
    language: English
    refInfo:
    copyright:
      @attributes:
        flag: Y
      custom: Language Resources & Evaluation is a copyright of Springer, 2024. All Rights Reserved.
      item: Language Resources & Evaluation
      holder: Springer Nature
      dt:
        @attributes:
          year: 2024
    holdings:
      @attributes:
        islocal: N