Corpus Alignment for Machine Translation | Pitch

Machine translation works really well these days, but it requires very high quality training data, in the form of source language - target language sentence pairs. To keep pace with recent progress, ModernMT, Gartner's 2019 only Cool Vendor for Machine Translation, sponsored Francesco Cariaggi and James McGuigan at School of AI 2020 Q4 to develop a novel algorithm to ensure the alignment of MT training texts. With the help of their mentor Marco Turchi, Head of Language Technology at Fondazione Bruno Kessler, they used language-agnostic sentence embeddings and dynamic programming. Are you an outstanding Engineer ready to make a difference in the AI field? Apply here: pischool.link/apply-students-school-of-ai Are you a Company facing a business challenge that can be solved by AI? Become a sponsor here: pischool.link/apply-sponsor-school-of-ai