A World Without English
Artūrs Vasiļevskis & Mārcis Pinnis
Thursday, 16 April, 2020
For many years, English dominated the digital world, but in the last couple years, demand has grown for the “other” 6,900+ languages. For the translation industry, using English as an intermediary language can be time-consuming, ineffective, expensive and, overall, a logistical nightmare. However, thanks to emerging AI technologies there is a viable new solution. With relatively small volumes of translated text, it is possible to quickly create useful AI driven machine translation solutions for unusual, tricky language combinations without resorting to first translating through English. In this webinar, Pinnis will reveal how AI-driven technologies can help keep up with increasing translation volumes and solve typical challenges seen in the translation industry. Vasilevskis will present two examples of how these technologies have been implemented in real translation scenarios and have created a paradigm shift in the traditional approach to problem solving in localization.
Arturs Vasilevskis is the Head of Machine Translation at Tilde, where he leads the Machine Translation group, overseeing all aspects of MT product development and sales. Under his leadership, Tilde has realized many major language technology projects for eGovernance, such as EU Council Presidency Translator, language technology platform hugo.lv and developed the Tilde MT solution for enterprises. His team has been recognized internationally through its victories in the World Machine Translation competition for three consecutive years (WMT2017, WMT2018 and WMT2019).
Doctor of Computer Science, Marcis Pinnis (M) is the Chief AI Officer at Tilde. Pinnis is an AI-powered technology expert, and his current research focus is neural machine translation (including hybrid methods for neural machine translation, terminology integration in neural machine translation, domain adaptation strategies of neural machine translation systems, etc.). While at Tilde, Pinnis has been involved in many EU funded projects, including FP7 ACCURAT, TaaS, TTC, and MultilingualWeb-LT projects, ICT PSP LetsMT! project, H2020 QT21, as well as several nationally funded projects in Latvia, Lithuania, and Estonia.
Most machine translation output is fine for basic Western LSP needs, as they usually translate within European...
Machine translation is great when it works, but it fails often and in unpredictable ways. What causes bad machine...