Discriminating Data: Neighborhoods, Individuals, Proxies


ACLS Fellowship Program


Modern Culture and Media


Discriminating Data—a monograph and series of digital projects—investigates the persistence and transformation of categories of race, gender, class, and sexuality in the era of network analytics. Analyzing the ways in which allegedly neutral machine learning algorithms merge methods to target individuals with those designed to group users into neighborhoods based on similar likes and dislikes, this project both reveals the assumptions underlying these algorithms and creates methods to desegregate them. By doing so, it produces a practice and theory of networked actions-as-speech that explores our potential as characters in a universe of unfolding dramas called “big data.”