Since its inception as an academic concept in the 1980s, intersectionality as a framework has traveled across various academic and social disciplines. Like many ideas, it has undergone adaptations in response to different environments, leading to deviations from its original intended meaning with each transfer. Computer sciences, particularly machine learning, have been heavily influenced by the introduction of intersectionality as both a theory and methodology. This influence is likely attributed to the increased exposure of intersectional bias instances. Our research seeks to examine the interpretation and application of intersectionality within the domain of machine learning. By analyzing three case studies through the lens of the theory’s initial proponents, we aim to provide comprehensive guidelines that address the challenge of evaluating the appropriate integration of intersectionality into machine learning practice.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.