As interest in neural machine translation reaches new heights — with even Mark Zuckerberg posting about it on his personal Facebook page — academia continues to churn out NMT papers at a record clip; 2017 output already on par with whole of 2016. Continue reading
Korea pits professional translators against assortment of neural machine translation engines. Contest shows that — unlike in Go — in terms of translation accuracy, humans remain firmly in the lead. Continue reading
Harvard and Systran publish a paper on their open-source neural machine translation model, OpenNMT. Research toolkit is a complete NMT implementation. Creators hope it will become the Moses of NMT. Continue reading
“Google, we promise we are not taking you on,” says the Harvard NLP group as a research project draws Systran’s attention and is co-developed into open-source NMT. Systran expects competitors to quickly build products based on the technology. Continue reading
Tokyo-based mini-conglomerate Fidel Technologies acquires India’s LinguaNext; rebrands to LinguaSol. It is Fidel’s second acquisition in 2016 following the purchase of a majority stake in a Japanese patent translation firm. Continue reading
Machine translation vendor opens demo version of its neural machine translation engine to the public; platform built on Torch. As with other NMT systems, model tends to improve fluency, but also produces new kinds of errors unseen in SMT. Continue reading
A roundup of recent appointments to director, business development, and senior manager posts at US Translation Company, Moravia, Akorbi, Systran, Choice Translating, SmartIdiom, TransPerfect, CSoft, MultiLing, TextMinded, Edimart, GPI Continue reading
This information will never be shared with third parties. No Spam.