The period of disappointment came to an end in 1977 with the development of the METEO System by the University of Montreal. METEO was created to translate weather forecasts from English to French and could translate 80,000 words a day, making it one of the first successful applications of the technology. The following year, Xerox began using Systran to translate its user manuals.
It has been a long time since machines have been limited to literal translations, complete with all the often comical mistakes. Who among us does not recall with tears of laughter that first version of Google Translate way back in 2006?
In this era of globalisation there is more translation work than ever. International businesses reach increasingly greater levels of sophistication, demanding and creating an enormous volume of documents and correspondence and a huge number of web pages, involving partners from all over the world with different cultures and languages.
The Twenty-first Annual Conference of the European Association for Machine Translation (EAMT) (EAMT 2018) will be held in Alacant/Alicante (Alacant in Catalan and Alicante in Castilian Spanish), Spain, from 28 to 30 May. The venue is a centuries-old building in the city centre.
Following Amazon’s announcement of its own neural machine translation (NMT) offering, the company’s machine learning scientists published a paper on research repository Arxiv.org that details the inner workings of Sockeye, their open-source, sequence-to-sequence toolkit for NMT.