Stay on top of how technology shapes the language services industries
Over 60 contributors in 15 African countries are gathering data and developing machine translation models to translate the Internet into indigenous languages. Continue reading
Team of Amazon AI researchers combines various deep-learning models to build automatic dubbing system. Viewer feedback reveals unnaturalness of speaking rate as main problem. Continue reading
As machine translation quality improves, precision-based metrics struggle to identify the best systems. Why have newer metrics not ousted longtime standard BLEU? Continue reading
Four college students harnessed recent Google Cloud releases to build a translation app that works offline and makes forms accessible to users with disabilities. Continue reading
As the language industry transitions to an operating model where the first draft is mostly generated by machines, many are looking for ways to quantify linguists' value-add. Edit-distance can help. Continue reading
An October 2019 paper by the Australian Strategic Policy Institute examines how China can employ machine translation tools used by consumers to collect data on users outside the country. Continue reading
Less than one year after Google open-sourced its much-discussed language model BERT, experts weigh in on its potential uses in neural machine translation. Continue reading
Mixed reactions from linguists and industry experts as search giant shelves its in-house translation productivity tool Translator Toolkit, which failed to gain real traction since launching in 2009. Continue reading
Leaked documents list additional machine translation capabilities for Siri as part of an upgrade for 2021; tech giant publishes rare neural MT research paper. Continue reading
Facebook has been granted a patent for its machine translation quality evaluation system, which relies on user engagement data to decide which translation users prefer. Continue reading
Baidu launches API for speech-to-speech translation based on technology outlined in late-2018 research paper, which was criticized at the time for overpromising. Continue reading
Characterizing post-editors based on their keystrokes, mouse activity, and other editing actions can help predict speed and, eventually, quality, says a new paper by Unbabel researchers. Continue reading
New research by Google AI explores multilingual NMT on an unprecedented scale. Good for low-resource languages, less so for the rest, results show. Continue reading
Venture capitalists see opportunity as startups begin to apply breakthroughs in transfer learning and new language models, such as BERT, ELMo, OpenAI's GPT-2, and ULMFiT. Continue reading
Virtual assistants are already part of the e-commerce setting, but taking them multilingual and training them for conversations with multiple variables still poses challenges. Continue reading
This information will never be shared with third parties. No Spam.