Digital social science

A relatively new topic in digital social science is the checking of computer algorithms for discrimination. Programs that use artificial intelligence and machine learning are just using data to learn patterns from them. This might reinforce easily observable stereotypes by matching information from different sources or identifying patterns in the a huge data set. Gender or ethnic discrimination is a frequently occurring nuisance. Sociologists know such types of misbehavior also as “statistical discrimination”. A new way of testing for such forms of discrimination in algorithms has been developed and can be studied and applied referring to an award-winning paper and software by Galhotra, Brun and Meliou. Take a closer look here:
https://github.com/LASER-UMASS/Themis?_ga=2.80913694.822913789.1503921109-1714816878.1503921076