Neural networks help discover how languages are related

Description

Historical linguistics is concerned with reconstructing the history and relatedness between languages of the world. Recently, this discipline gained new insights by applying computational methods, such as new proposals for the birthplace of the Indo-European language family, to which most European languages belong (Bouckaert et al., 2012). As I am passionate about both historical linguistics and machine learning, I decided to combine these two fields in my master’s thesis at the University of Amsterdam CLC lab in 2018. In this blog post at the website of the CLC lab, I describe my master's thesis research and its implications.