scispace - formally typeset
M

Marine Carpuat

Researcher at University of Maryland, College Park

Publications -  139
Citations -  3250

Marine Carpuat is an academic researcher from University of Maryland, College Park. The author has contributed to research in topics: Machine translation & Computer science. The author has an hindex of 28, co-authored 124 publications receiving 2463 citations. Previous affiliations of Marine Carpuat include National Research Council & Columbia University.

Papers
More filters
Journal ArticleDOI

BLOOM: A 176B-Parameter Open-Access Multilingual Language Model

Teven Le Scao, +386 more
- 09 Nov 2022 - 
TL;DR: BLOOM as discussed by the authors is a decoder-only Transformer language model that was trained on the ROOTS corpus, a dataset comprising hundreds of sources in 46 natural and 13 programming languages (59 in total).
Proceedings Article

Improving Statistical Machine Translation Using Word Sense Disambiguation

Marine Carpuat, +1 more
TL;DR: This paper investigates a new strategy for integrating WSD into an SMT system, that performs fully phrasal multi-word disambiguation, and provides the first known empirical evidence that lexical semantics are indeed useful for SMT, despite claims to the contrary.
Proceedings ArticleDOI

Word Sense Disambiguation vs. Statistical Machine Translation

Marine Carpuat, +1 more
TL;DR: It is found that word sense disambiguation does not yield significantly better translation quality than the statistical machine translation system alone.
Proceedings Article

Task-based Evaluation of Multiword Expressions: a Pilot Study in Statistical Machine Translation

Marine Carpuat, +1 more
TL;DR: Two different integration strategies for MWE inSMT are proposed, which take advantage of different degrees of MWE semantic compositionality and yield complementary improvements in SMT quality on a large-scale translation task.
Posted Content

An Empirical Exploration of Curriculum Learning for Neural Machine Translation

Xuan Zhang, +8 more
- 02 Nov 2018 - 
TL;DR: A probabilistic view of curriculum learning is adopted, which lets us flexibly evaluate the impact of curricula design, and an extensive exploration on a German-English translation task shows it is possible to improve convergence time at no loss in translation quality.