For a Human-Centered AI

EAMT Best Thesis Award 2016 assigned to José Camargo de Souza!

June 8, 2017

The prize is awarded annually by the European Association for Machine Translation (MT) for the best thesis on a topic related to MT and, for the 1st time, it has been won by a student from our labs.

Another acknowledgment of the quality of research and education at FBK: our former PhD student in the HLT-MT unit won the EAMT Best Thesis Award 2016.

José’s thesis on “Adaptive Quality Estimation for Machine Translation and Automatic Speech Recognition” has been selected among 7 finalists by an international panel of experts that this year evaluated a record number of submissions from all the top European universities.

Few words from his advisors:

Marcello Federico: “This award is the result of advanced research in the HLT-MT unit on a wide range of MT-related and application-oriented problems. José’s work on the automatic prediction of MT output quality is an important step towards bypassing the need of manually created “reference” translations (a major cost and bottleneck in current evaluation protocols) and integrate quality predictions in industrial translation workflows as a way to increase human productivity”.

Matteo Negri: “José ability and his commitment to this difficult problem allowed him to follow the ideal path for a PhD career. Hard study, attention to networking, release of open-source code and successful internships abroad were just some of the ingredients of a recipe that eventually led to a strong record of publications and, most importantly, to José’s current position as research scientist in the translation team at eBay in Aachen, Germany”.

Marco Turchi: “This important award is just the cherry on the cake. Talking about predictions, it’s easy to predict a brilliant career for José!”.