Reclaiming Lost Data on American Racial Inequality - Incite at Columbia University

We produced a big-data genealogy of the African-American past by combining algorithmic linking techniques with historical and genealogical methods.

To do so, we drew on the development of machine-learning algorithms to link individual census records over time with idiosyncratic data sources such as letters, marriage records, church registries, and oral histories.

Support from the Russell Sage Foundation allowed us to develop new methods for linking historical data using rarely consulted types of historical evidence. This work is rooted in the idea that in order to link marginalized groups often excluded or missed from official tabulations, we need to rely on additional sources of historical and genealogical information.

Incite is collaborated with researchers at the University of California-Berkeley, Harvard University, the Ohio State University, and the University of Washington to carry out this work.