NamSor has been extensively used by H.R. platforms to provide recruiters with innovative diversity dashboards: is there a gender, racial or ethnic bias in some part of a recruitment process? In Q3-Q4 2019, we plan to open a new API to evaluate gender, ethnic and racial biases in any machine learning process that follow a ‘funnel’ model (such as : recruitment, credit allocation, banking account opening and KYC etc.).
Zack Kertcher, principal of Research Done, has conducted innovative research on the racial breakdown trends in some US public services, and the results have been published in a WBEZ article :
Part of the data analysis project included comparing NamSor’s precision and recall rates with alternatives. Open source classifiers (such as : appeler / ethnicolr) provide functionalities that are similar. Our commitment at NamSor is to try and provide a variety of taxonomies that reveal different aspects of identity through names, the best possible accuracy and a global coverage (all countries, all regions, all alphabets).
NamSor™ Applied Onomastics is a European vendor of sociolinguistics software (NamSor sorts names).
Image credits : Elliott Ramos/WBEZ