Author: Kelsey Campbell
Publisher: Gay TA Science
Publication Year: 2021
Summary: In the following talk, Kelsey explains how the cissexism present in society today impacts data science as a field. They give examples from real-life on cissexism’s impact. For example, facial recognition technology has caused several issues for trans people. A trans person’s appearance may change over time, so an algorithm that relies on a static picture of someone to recognize them may have trouble identifying a trans person as themselves. Also, when facial recognition technology tries to guess someone’s gender, the algorithms behind it usually do not consider gender beyond what a cisgender man looks like versus what a cisgender woman looks like. In order to combat the cissexism present in algorithms, Kelsey suggests that data scientists who work with algorithms where gender impacts the results actively consider the potential for bias throughout the data collection and model-building processes. Once the potential for bias has been considered, the data collection processes and algorithms should be modified to eliminate the bias.
