MARGARET: Multidimensional Algorithm Generated Anthropological Recording and Ethnographic Tool
Margaret is a Disruptioneering Project under DARPA’s Understanding Group Bias Program, with the goal to quantify social and cultural conditions that impact gender disparities within cultural groups. The goal of the project is to develop ML algorithms that are able to quantify gender biases inherent in data and help explain parameters that influence gender equality. Further, we investigate the relationship between explicit cultural data (e.g., gender inequality indices) and implicit gender biases (representative of attitudes, beliefs), using “thick” i.e. real-world recordings of subjective experience (“from the ground” recordings) as well as “big” data.