Literary Canonicity and Algorithmic Fairness: The Effect of Author Gender on Classification Models

TY - GEN

T1 - Literary Canonicity and Algorithmic Fairness

T2 - 2024 Computational Humanities Research Conference, CHR 2024

AU - Lassen, Ida Marie S.

AU - Feldkamp, Pascale

AU - Bizzoni, Yuri

AU - Nielbo, Kristoffer Laigaard

N1 - Conference code: 5

PY - 2024

Y1 - 2024

N2 - This study examines gender biases in machine learning models that predict literary canonicity. Using algorithmic fairness metrics like equality of opportunity, equalised odds, and calibration within groups, we show that models violate the fairness metrics, especially by misclassifying non-canonical books by men as canonical. Feature importance analysis shows that text-intrinsic differences between books by men and women authors contribute to these biases. Men have historically dominated canonical literature, which may bias models towards associating men-authored writing styles with literary canonicity. Our study highlights how these biased models can lead to skewed interpretations of literary history and canonicity, potentially reinforcing and perpetuating existing gender disparities in our understanding of literature. This underscores the need to integrate algorithmic fairness in computational literary studies and digital humanities more broadly to foster equitable computational practices.

AB - This study examines gender biases in machine learning models that predict literary canonicity. Using algorithmic fairness metrics like equality of opportunity, equalised odds, and calibration within groups, we show that models violate the fairness metrics, especially by misclassifying non-canonical books by men as canonical. Feature importance analysis shows that text-intrinsic differences between books by men and women authors contribute to these biases. Men have historically dominated canonical literature, which may bias models towards associating men-authored writing styles with literary canonicity. Our study highlights how these biased models can lead to skewed interpretations of literary history and canonicity, potentially reinforcing and perpetuating existing gender disparities in our understanding of literature. This underscores the need to integrate algorithmic fairness in computational literary studies and digital humanities more broadly to foster equitable computational practices.

KW - algorithmic fairness

KW - bias

KW - canonicity

KW - computational literary studies

KW - gender bias

UR - https://www.scopus.com/pages/publications/85210828063

M3 - Conference article

SN - 1613-0073

VL - 3834

SP - 153

EP - 171

JO - CEUR Workshop Proceedings

JF - CEUR Workshop Proceedings

M1 - 76

Y2 - 4 December 2024 through 6 December 2024

ER -