All my students know that I’m a big fan of the online dictionary Linguee. It’s great for looking up words and even phrases. But if you wanted to see a whole paragraph translated you would need to resort to Google translate, as Linguee often even can’t cope with longer sentences ….. until now!
A few days ago I found out that Linguee has gone a step further by developing a machine translator to rival google: DeepL Translator
So I went to see how well it works. First I chose a German text and used google translate to translate it:
I then used DeepL translator:
Although Google didn’t give a bad translation, I found DeepL to be more accurate. It also offered you the option of checking out individual words from the original text. I clicked on ‘erforderlich’ and received the familiar Linguee translations of the word:
There is also the option to check for alternative translations (similar to Google). I clicked on the word ‘renovated’ and got the following choices, again with the familiar Linguee dictionary.
I was very impressed, so I tried it with an English text. Google translate again first:
Again, not bad, although I wouldn’t have chose the words ‘Sorgen gemacht’ for ‘concerns raised’ or ‘Einbruch’ for slump. Could DeepL do it better?
Yes, the choice of ‘Besorgnisse geäußert’ for ‘concerns raised’ and ‘Rückgang’ for slump was in this case far better. I still clicked on the word ‘concerns’ to see what other options might be available:
It showed me that ‘Rückgang’ was a good choice for decline, although instead of ‘Besorgnisse’ for concerns I might have chosen ‘Bedenken’.
All in all an impressive translating machine! But remember, it is a machine! There may – I think I can safely say there will – be mistakes!
It should, however, give you a better idea of an unknown text if you come across a foreign website where something looks interesting but you can’t read it. It should also be a great help for translators, especially when it comes to keywords. Just remember that you shouldn’t rely on the translation.
I’ll add the press release that I received from them for those who might be interested.
Anybody else, go to DeepL Translator, try it out, then come back and tell me what you think!
Cologne, Germany, August 29th, 2017
German tech company DeepL launches DeepL Translator, a breakthrough in neural machine translation that delivers translations of unrivaled quality. From today, it is available to everyone, free of charge on DeepL.com.Despite the money and resources that major internet companies have been pumping into the development of neural networks, it is DeepL that can boast the world’s most accurate and natural-sounding machine translation tool. When users enter a text, DeepL’s artificial intelligence is able to capture even the slightest nuances and reproduce them in translation unlike any other service.
In blind tests pitting DeepL Translator against the competition, translators preferred DeepL’s results by a factor of 3:1. Automated tests bear this out, as well. Within the field of machine translation, the gold standard for measuring system performance is the bilingual evaluation understudy (BLEU) score, which compares machine-translated texts with those produced by a translator. DeepL Translator achieves record BLEU scores.*
“We have achieved several significant improvements in neural network architecture,” says Gereon Frahling, the company’s founder and CEO. “By arranging the neurons and their connections differently, we have enabled our networks to map natural language more comprehensively than any other neural network to date.”
DeepL’s revolutionary neural architecture runs on a supercomputer in Iceland, capable of 5.1 petaFLOPS (5 100 000 000 000 000 operations per second), enough power to translate a million words in under a second. “Due to the abundance of renewable energy, we can train our neural networks very cost-efficiently in Iceland. We will continue to invest in high-performance hardware,” explains CTO Jaroslaw Kutylowski.
The DeepL team uses this supercomputer to train neural translation networks on a huge collection of multilingual texts. In training, the networks examine a multitude of translations and learn independently how to translate with correct grammar and structure. Here, DeepL can draw on the success of its first product: Linguee, the world&rsqo;s largest translation search engine. Over the past decade, DeepL has gathered over a billion high-quality translated texts, the finest possible training material for a neural translation network.
DeepL Translator currently supports 42 language combinations between English, German, French, Spanish, Italian, Polish, and Dutch. The neural networks are already training to master more languages like Mandarin, Japanese, and Russian. DeepL also intends to release an API in the coming months, allowing its superior translations to enhance other products such as digital assistants, dictionaries, language learning apps, and professional translation programs.
“DeepL’s ambitions are not limited to translation,” says Frahling. “Our neural networks have developed a level of text understanding that opens several exciting possibilities.”
* The previous record on the widely-used WMT 2014 newstest set for English-German was 28.4, achieved by the Google research department. On the same set, DeepL’s public translation service obtains a BLEU score of 31.1. For the 2014 newstest set for English-French, DeepL Translator achieves a BLEU score of 44.7, besting the previous record by more than 3 points.
DeepL is a deep learning company that develops AI systems for languages. The company was founded in 2009 as Linguee and introduced the first internet search engine for translations. Linguee has answered over 10 billion queries from more than 1 billion users. It is one of the top 200 most widely-used websites in France, Germany, Spain, and other countries.
In 2017, the company introduced DeepL Translator, a machine translation system that is able to produce translations of unprecedented quality, available for free at DeepL.com/translator.
DeepL continues to develop artificial intelligence products that will expand human possibility, overcome language barriers, and bring cultures closer together.
Im Mediapark 8a
Register Court Köln, HRB 64604