#Latest stories for ZDNet related to EU * Edition: + Asia + Australia + Europe + India + United Kingdom + United States + ZDNet around the globe: + ZDNet China + ZDNet France + ZDNet Germany + ZDNet Korea + ZDNet Japan Search ____________________ (BUTTON) Go * Central Europe * Middle East * Scandinavia * Africa * UK * Italy * Spain * more + Estonia + European Union + Mobility + Hardware + Innovation + See All Topics + White Papers + Downloads + Reviews + Galleries + Videos + TechRepublic Forums * Newsletters * All Writers * + Log In to ZDNET + Join ZDNet + About ZDNet + Preferences + Community + Newsletters + Log Out * * + ____________________ (BUTTON) Go * Menu + Central Europe + Middle East + Scandinavia + Africa + UK + Italy + Spain + Estonia + European Union + Mobility + Hardware + Innovation + See All Topics + White Papers + Downloads + Reviews + Galleries + Videos + TechRepublic Forums * + o Log In to ZDNET o Join ZDNet o About ZDNet o Preferences o Community o Newsletters o Log Out * eu + Asia + Australia + Europe + India + United Kingdom + United States + ZDNet around the globe: + ZDNet China + ZDNet France + ZDNet Germany + ZDNet Korea + ZDNet Japan Google announces Neural Machine Translation to improve Google Translate The Google Neural Machine Translation system 'surpasses' the results of all other machine-translation solutions currently available, with GNMT now being used for Chinese-to-English translations. Corinne Reichert By Corinne Reichert | September 28, 2016 -- 05:04 GMT (06:04 BST) | Topic: Artificial Intelligence * 0 * * * * Google has announced a neural machine translation (NMT) system that it says will reduce translation errors across its Google Translate service by between 55 percent and 85 percent, calling the achievement by its team of scientists a "significant milestone". Announced on the 10-year anniversary of Google Translate, Google Brain Team research scientists Quoc Le and Mike Schuster said Google has used machine intelligence to improve its image and speech-recognition systems, but had previously found it "challenging" to improve machine translation until now. Featured stories * Best of CES 2019: Cool tech you can buy this year * Game of Clouds: Lock-In is Coming * Samsung's Galaxy S10 unveiling is February 20 and all it has to do is save the smartphone industry * How to build a temporary green screen YouTube studio "Ten years ago, we announced the launch of Google Translate, together with the use of Phrase-Based Machine Translation as the key algorithm behind this service," they said. "Today, we announce the Google Neural Machine Translation system (GNMT), which utilizes state-of-the-art training techniques to achieve the largest improvements to date for machine translation quality." Unlike the currently used phrase-based machine translation (PBMT) system -- which translates words and phrases independently within a sentence, and is notorious for its mistranslations -- neural machine translation considers the entire sentence as one unit to be translated. Researchers have been working on improving neural machine translation over the past few years, with Google's scientists finding a way to make it work on large data sets while maintaining speed and accuracy. Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation [PDF], by Quoc Le, Mike Schuster, Yonghui Wu, Zhifeng Chen, and Mohammad Norouzi et al, said the system is "state of the art" for English-to-French and English-to-German translations in particular, reducing errors by 60 percent on average. "Our model consists of a deep LSTM [long short-term memory] network with eight encoder and eight decoder layers using residual connections as well as attention connections from the decoder network to the encoder. To improve parallelism and therefore decrease training time, our attention mechanism connects the bottom layer of the decoder to the top layer of the encoder," the technical report said. "To accelerate the final translation speed, we employ low-precision arithmetic during inference computations. To improve handling of rare words, we divide words into a limited set of common sub-word units ('wordpieces') for both input and output. This method provides a good balance between the flexibility of 'character'-delimited models and the efficiency of 'word'-delimited models, naturally handles translation of rare words, and ultimately improves the overall accuracy of the system." Google used the example of a Chinese sentence being translated into English: google-neural-machine-translation-system-2.png (Image: Google) "First, the network encodes the Chinese words as a list of vectors, where each vector represents the meaning of all words read so far ('Encoder'). Once the entire sentence is read, the decoder begins, generating the English sentence one word at a time ('Decoder'). To generate the translated word at each step, the decoder pays attention to a weighted distribution over the encoded Chinese vectors most relevant to generate the English word," Google said. "Using human-rated side-by-side comparison as a metric, the GNMT system produces translations that are vastly improved compared to the previous phrase-based production system." Google has launched GNMT in production across Google Translate on web and mobile for the Chinese-to-English translation pair, accounting for around 18 million translations daily. It will roll out GNMT with more languages over the coming months. google-neural-machine-translation-system.png google-neural-machine-translation-system.png (Image: Google) "Our key findings are: That wordpiece modeling effectively handles open vocabularies and the challenge of morphologically rich languages for translation quality and inference speed; that a combination of model and data parallelism can be used to efficiently train state-of-the-art sequence-to-sequence NMT models in roughly a week; that model quantization drastically accelerates translation inference, allowing the use of these large models in a deployed production environment; and that many additional details like length-normalization, coverage penalties, and similar are essential to making NMT systems work well on real data," the technical report concluded. Google used its machine-learning toolkit TensorFlow and its Tensor Processing Unit (TPU) chips to bring GNMT into production with sufficient latency and power. At Google I/O in May, Google said machine learning and artificial intelligence will be its big bet into the future of the business. "We're working hard at core use cases on mobile," said CEO Sundar Pichai during the Google I/O keynote. "We are pushing ourselves really hard so Google is staying a step ahead of our customers." Staying ahead in mobile, according to Pichai, means providing a better, more assisted experience by integrating machine-learning tools through its cloud platform and open-source community via TensorFlow. "We can do things never could do before," said Pichai. "We believe we're at a seminal moment in next 10 years and want to take next step to be more assistive. We want to be there for our users. We want to help you get things done in the real world and understanding your context. "The real test is whether humans can achieve a lot more with AI assisting them. We look forward to building this future." Related Topics: Google Digital Transformation CXO Internet of Things Innovation Enterprise Software * 0 * * * * LOG IN TO COMMENT * My Profile * Log Out | Community Guidelines Join Discussion Add Your Comment Add Your Comment More from Corinne Reichert * Mobility Ericsson and Deutsche Telekom hit 40Gbps wireless backhaul speeds * 5G CES 2019: The biggest 5G news * 5G Sprint rounds out CES 2019 with 5G call * Innovation CES 2019: Intel details autonomous vehicle trial in Israel × Thank You Please review our terms of service to complete your newsletter subscription. By registering you become a member of the CBS Interactive family of sites and you have read and agree to the Terms of Use, Privacy Policy and Video Services Policy. You agree to receive updates, alerts and promotions from CBS and that CBS may share information about you with our marketing partners so that they may contact you by email or otherwise about their products or services. You will also receive a complimentary subscription to the ZDNet's Tech Update Today and ZDNet Announcement newsletters. You may unsubscribe from these newsletters at any time. (BUTTON) ACCEPT & CLOSE Newsletters See All See All Related Stories * 1 of 3 * * CES 2019: The biggest 5G news The biggest 5G announcements out of CES 2019 included Samsung's prototype 5G smartphone, Intel's SoC, Verizon's 5G showcase with Disney, Sprint's 5G and IoT combination, Qualcomm's ... * CES 2019: Qualcomm President Amon is convinced you’re going to be thrilled with 5G Qualcomm's president, Cristiano Amon, took some time at the Consumer Electronics Show to discuss how 5G cellular will amaze consumers. He also reflected on how AI processing on smartphones ... * Sprint rounds out CES 2019 with 5G call Working with Nokia and Qualcomm, Sprint has made a 5G data call on its live commercial network in San Diego, including streaming YouTube videos, making Skype video calls, and sending ... * CES 2019: Intel details autonomous vehicle trial in Israel Intel is extending its autonomous driving trials in Israel to provide an entire mobility-as-a-service offering across car, software, platform, mapping, and safety mechanisms. ... * CES 2019: AI startup Eyeris will know your car from the inside-out Cars on the road know little about drivers and passengers today -- except maybe whether the person at the wheel is drowsy. Tomorrow's vehicle will have a total awareness of the cabin, ... * Microsoft's LinkedIn report card: Technology integration continues, but at a slow pace Microsoft and its LinkedIn subsidiary are continuing to integrate on the productivity and AI fronts in a deliberately slow and incremental way, based on a look at their most recent ... * CES 2019: Cortana's gone underground, but Microsoft is not out of the virtual assistant game Cortana was a no-show at CES 2019 this week. But Microsoft's Virtual Assistant Accelerator, which lets customers build their own personalized digital assistants, racked up a few ... * CES 2019: Cisco talks 6G While everyone else spent CES 2019 talking about 5G, Cisco is already looking towards a 6G future. * CES 2019: The Trade Desk CEO says AI helps humans, improves business Jeff Green, chief executive of advertising industry giant The Trade Desk, says a thoughtful use of AI can free up people do to what humans do best: be creative and connect with other ... ZDNet Connect with us © 2019 CBS Interactive. All rights reserved. Privacy Policy | Cookies | Ad Choice | Advertise | Terms of Use | Mobile User Agreement Visit other CBS Interactive sites: [Select Site_____] * Topics * All Authors * Galleries * Videos * Sponsored Narratives * About ZDNet * Meet The Team * Site Map * RSS Feeds * Reprint Policy * Manage | Log Out * Join | Log In | Membership * Newsletters * Site Assistance * ZDNet Academy * TechRepublic Forums