Tech Pioneer: Maja Popović is one of ADAPT’s Innovating Women
Highlighting ADAPT’s pioneering technologists, we celebrate Research Fellow Maja Popović.
An expert in Machine Translation (MT), Maja Popović has extensive experience in evaluation of machine translation which is a complex task and represents a research sub-field itself. She is also the author of the automatic metric ‘chrF’ based on matching between character sequences which has been widely accepted in the MT community and also used for other NLP tasks. It also became one of the official metrics in the WMT shared task, the most important MT evaluation campaign in the field. The original metric is publicly available as an open source Python script, together with several additional versions published by other researchers and users.
Widely published in peer reviewed journals, Maja’s research interests include machine translation, multilingual NLP, human and automatic evaluation methods for NLP. She has also worked on several industry projects with Microsoft, Linkedin, and Optum, dealing with machine translation for different domains, reducing gender bias in NLP, and multilingual text classification.