@inproceedings{koudounas-etal-2025-privacy,
    title = "Privacy Preserving Data Selection for Bias Mitigation in Speech Models",
    author = "Koudounas, Alkis  and
      Pastor, Eliana  and
      Mazzia, Vittorio  and
      Giollo, Manuel  and
      Gueudre, Thomas  and
      Reale, Elisa  and
      Cagliero, Luca  and
      Cumani, Sandro  and
      De Alfaro, Luca  and
      Baralis, Elena  and
      Amberti, Daniele",
    editor = "Rehm, Georg  and
      Li, Yunyao",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.acl-industry.52/",
    pages = "738--748",
    ISBN = "979-8-89176-288-6",
    abstract = "Effectively selecting data from subgroups where a model performs poorly is crucial for improving its performance. Traditional methods for identifying these subgroups often rely on sensitive information, raising privacy issues. Additionally, gathering such information at runtime might be impractical. This paper introduces a cost-effective strategy that addresses these concerns. We identify underperforming subgroups and train a model to predict if an utterance belongs to these subgroups without needing sensitive information. This model helps mitigate bias by selecting and adding new data, which is labeled as challenging, for re-training the speech model. Experimental results on intent classification and automatic speech recognition tasks show the effectiveness of our approach in reducing biases and enhancing performance, with improvements in reducing error rates of up to 39{\%} for FSC, 16{\%} for ITALIC, and 22{\%} for LibriSpeech."
}