TQ-AutoTest: Novel analytical quality measure confirms that DeepL is better than Google Translate
Vivien Macketanz, Aljoscha Burchardt & Hans Uszkoreit
Monday, 17 December, 2018
In the second half of 2017 news broke that DeepL, a Machine Translation system built by a small German company, was able to beat Google and other known systems in terms of translation quality (see, e.g., this article in TechCrunch). Using the new semi-automatic tool TQAutoTest that allows for an informative, analytical comparison of different MT engines, the Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI) could confirm this observation for the language pair German – English they examined.
Vivien Macketanz, M.A., is a researcher at Language Technology Lab of the German Research Center for Artificial Intelligence (DFKI). Before joining DFKI in 2016, she studied linguistics and communication at the University of Potsdam and Technical University Berlin. Her main research focus lies on the linguistic evaluation of Machine Translation. She is currently working on her PhD for which she is developing a scheme for semi-automatic checking of Machine Translation output.
Aljoscha Burchardt works in the Language Technology Lab of the German Research Center for Artificial Intelligence (DFKI GmbH). He manages the EC-funded project QTLaunchPad that is preparing a big European translation quality initiative. He is also a manager within the European Network of Excellence META-NET that is preparing a Technology Alliance for Multilingual Europe and project leader of the taraXÜ project that develops hybrid machine translation technology in a consortium with industry partners. Burchardt has a background in semantic Language Technology. After his PhD in Computational Linguistics at Saarland University he coordinated the Center of Research Excellence "E-Learning 2.0" at Technische Universität Darmstadt.
Prof. Dr. Hans Uszkoreit is Scientific Director at the German Research Center for Artificial Intelligence (DFKI). He has been doing research in Artificial Intelligence for more than 30 years. He is leading research in different research institutes, universities and industry in the USA, Germany and China. For 20 years, he was a Professor of Computational Linguistics and Computer Science at the Saarland University and since 2011 he is Honorary Professor at the Technical University of Berlin. Since 2017 he additionally leads the Artificial Intelligence Technology Center in Beijing and is furthermore Chief Technology Advisor of the Lenovo Corporation. His main interest lies in the foundation and application of language and knowledge technology. He is author and co-author of more than 200 international publications.
This presentation will discuss of some of the challenges that make pricing difficult once MT is in play. Using some...
How to do that, though, is complicated by several factors:
Resistance to lower post-edit fees for MT by qualified...
Nikon Precision started utilizing machine translation technology in its localization workflows a little over 5 years...