Discriminating Data: Neighbourhoods, Individuals and Proxies

This project investigates the centrality of race, gender, class and sexuality to big data and network analytics. Unpacking key technical concepts—from correlation to proxies, factor analysis to deep learning—it reveals how these principles often foster acrimony and segregation through their default assumptions and conditions, defaults that amplify the societal and human prejudices that they were developed to combat. Discriminating Data argues for the imperative to develop alternative algorithms, defaults and interdisciplinary coalitions in order to desegregate networks and buttress social justice.