Identifying and Reducing Bias in Artificial Datasets
Was this section helpful?
Semantics derived automatically from language corpora contain human-like biases, Aylin Caliskan, Joanna J. Bryson, Arvind Narayanan, 2017Science, Vol. 356 (American Association for the Advancement of Science)DOI: 10.1126/science.aal4230 - This seminal paper demonstrates how word embeddings learned from text corpora capture societal biases, introducing the Word Embedding Association Test (WEAT), highly relevant to the 'Embedding Analysis' section.
Fairness and Machine Learning: Limitations and Opportunities, Solon Barocas, Moritz Hardt, Arvind Narayanan, 2023 (MIT Press) - An open-access book that provides a comprehensive look at the ethical considerations, definitions of fairness, and technical approaches to addressing bias in machine learning, offering a foundational understanding for the 'A Note on "Fairness"' section.