Skip to main content

CMU Portugal alumnus and IT researcher wins ERC Studentship

André Martins

André Martins, researcher at the Telecommunications Institute and alumnus of the CMU Portugal Program, has won a €2 millionConsolidator Studentship from the European Research Council (ERC) to study artificial neural networks applied to natural language processing (NLP).

André Martins was the first alumnus of the CMU Portugal Dual PhD program, having simultaneously obtained a PhD in Language Technologies from Instituto Superior Técnico and Carnegie Mellon University (CMU) in 2012. Since then, he has remained involved with CMU Portugal Program initiatives, notably leading two CMU Portugal projects – MAIA and GoLocal. He is also Vice President of Artificial Intelligence Research at Unbabel, an industrial affiliate of the CMU Portugal Program and a partner in several CMU Portugal Program research projects. An Associate Professor at Instituto Superior Técnico, André Martins is also co-founder and co-organizer of the Lisbon Machine Learning School (LxMLS), which is supported by the CMU Portugal Program.

Studentships Consolidator Studentships are intended to support researchers of any nationality with 7 to 12 years of experience since completing their PhD. This Studentship support the development of André Martins' DECOLLAGE (DEep COgnition Learning for LANguage GEneration) project, which will seek solutions to some fundamental problems in NLP, using an innovative interdisciplinary methodology that brings together artificial intelligence tools, sparse modeling, neuroscience, and cognitive sciences. According to the researcher, "this project is another step towards overcoming the limitations of current NLP technologies, enabling humans and machines to communicate effectively in natural language and work collaboratively to solve increasingly difficult problems."

This ERC Studentship follows on from an ERC Starting Grant awarded in 2017, worth €1.4 million, for the DeepSPIN project, which enabled research into structured statistical learning methods combined with artificial neural networks, applying them to natural language processing, including machine translation.