알고리즘 공정성의 실제와 사회과학의 역할 = The Practices of Algorithmic Fairness and the Role of Social Sciences

다국어 초록 (Multilingual Abstract)

As the rapid advancement of artificial intelligence has led to the widespread applications of algorithms in the society, the question of how to incorporate normative values, especially fairness criteria, into automated decision making has become an important agenda in Artificial Intelligence (AI) governance. The primary objective of this paper is to elucidate the key concepts and fundamental issues of algorithmic fairness by presenting examples that are accessible to readers with basic statistical knowledge. I address the pipeline of algorithm-assisted decision making, types of fairness concepts and their relationships, and approaches for embedding normative standards into the automated decision making pipeline. Finally, I explore potential research areas in related topics for social scientists and suggest that social sciences should progress from retrospective disciplines to practical disciplines.

참고문헌 (Reference)

1 Li, Bo, "Trustworthy AI: From Principles to Practices" 55 (55): 177-, 2023

2 Jobin, Anna, "The Global Landscape of AI Ethics Guidelines" 1 (1): 389-399, 2019

3 Lundberg, Ian, "The Gap-Closing Estimand: A Causal Approach to Study Interventions that Close Disparities across Social Categories" 53 (53): 507-570, 2024

4 Dressel, Julia, "The Accuracy, Fairness, and Limits of Predicting Recidivism" 4 (4): 2018

5 Keskintürk, Turgut, "Sociology’s Inequality Problem"

6 Council of the European Union, "Proposal for a Regulation of the European Parliament and of the Council Laying down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act) and Amending Certain Union Legislative Acts - Analysis of the Final Compromise Text with a View to Agreement"

7 Imai, Kosuke, "Principal Fairness for Human and Algorithmic Decision-Making" 38 (38): 317-328, 2023

8 Kleinberg, Jon, "Prediction Policy Problems" 105 (105): 491-495, 2015

9 Bishop, Christopher M, "Pattern Recognition and Machine Learning" Springer 2006

10 Kleinberg, Jon, "Inherent Trade-Offs in the Fair Determination of Risk Scores" 67 : 43:1-43:23, 2017

11 Coleman, James S, "Inequality, Sociology, and Moral Philosophy" 80 (80): 739-764, 1974

12 Larrazabal, Agostina J., "Gender Imbalance in Medical Imaging Datasets Produces Biased Classifiers for Computer-Aided Diagnosis" 117 (117): 12592-12594, 2020

13 Caton, Simon, "Fairness in Machine Learning: A Survey" 56 (56): 166-, 2024

14 Barocas, Solon, "Fairness and Machine Learning : Limitations and Opportunities" MIT Press 2023

15 Dwork, Cynthia, "Fairness Through Awareness" 214 (214): 214-226, 2012

16 Starke, C., "Fairness Perceptions of Algorithmic Decision-Making: A Systematic Review of the Empirical Literature" 9 (9): 2022

17 Berk, Richard A., "Fair Risk Algorithms" 10 : 165-187, 2023

18 Liao, Shu-Hsien, "Expert System Methodologies and Applications - A Decade Review From 1995 to 2004" 28 (28): 93-103, 2005

19 Obermeyer, Ziad, "Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations" 366 (366): 447-453, 2019

20 Chohlas-Wood, Alex, "Designing Equitable Algorithms" 3 (3): 601-610, 2023

21 Hurley, Mikella, "Credit Scoring in the Era of Big Data" 18 : 148-216, 2016

22 The New York Times, "Biased Algorithms Are Easier to Fix than Biased People"

23 Stanford University, "Artificial Intelligence Index Report 2024"

24 Rambachan, Ashesh, "An Economic Perspective on Algorithmic Fairness" 110 : 91-95, 2020

25 Massey, Douglas S., "American Apartheid: Segregation and the Making of the Underclass" Harvard University Press 1993

26 Das, Sanjiv, "Algorithmic Fairness" 15 : 565-593, 2023

27 Kleinberg, Jon, "Algorithmic Fairness" 108 : 22-27, 2018

28 Rawls, John, "A Theory of Justice" Harvard University Press 1971

29 Jasso, Guillermina, "A New Theory of Distributive Justice" 45 (45): 3-32, 1980