We are constantly asked whether we use certain tools or programs for our work at Mimikama.
The answer to this is usually not only quite astonishing for those asking, but actually helpful, because at Mimikama we don't use unicorn dust or magic wands. We use freely accessible tools and websites that anyone can work with. These tools shouldn't be a secret, so we're happy to introduce some of these useful services.
DeepL
For translations, primarily from English, we at Mimikama use DeepL ( www.DeepL.com ). We see the clear strengths of DeepL in the sensibly designed translations, i.e. DeepL “understands” texts and does not simply reproduce clumsy word translations, but rather an understandable presentation.
DeepL usually also recognizes when it involves technical terms or when there are proper names in the original text that do not require translation. It also recognizes phrases that, for example, the Google or Bing translator translates literally and therefore not so beautifully. As an example, one can certainly cite the cats and dogs metaphor, which in English describes a heavy rain, but does not work that way in German. It's not raining cats and dogs, it's raining heavily.
DeepL reacts much more elegantly here and therefore the translation is correspondingly more understandable. In addition to English, DeepL also translates French, Spanish, Italian, Dutch and Polish.
About DeepL
From the press release dated August 29, 2017.
German technology company DeepL today launches DeepL Translator, a new machine translation system that delivers translations of unprecedented quality. The DeepL Translator can be used www.DeepL.com
Although large American Internet companies invest immense amounts of money and resources in the development of neural networks for translation, the Cologne company DeepL is setting a new standard in machine translation. DeepL Translator is a website where users can have any text translated. The artificial intelligence understands even the smallest details and reflects them meaningfully in the translation.
In blind tests comparing DeepL Translator with well-known machine translations, professional translators rated DeepL's results as better three times more often. Automated tests also prove this. In the field of machine translation, the gold standard for measuring quality is the BLEU score, which compares machine-translated texts with those of a translator. The DeepL translator also achieves record results in terms of the BLEU score. *
“We have made some significant improvements to the architecture of the neural networks,” says Gereon Frahling, founder and managing director of the company. “By reorganizing the neurons and their connections, we have enabled our networks to represent natural language better than any previous neural translation network.”
DeepL's artificial intelligence runs on a supercomputer in Iceland that can perform 5.1 petaFLOPS (5,100,000,000,000,000 operations per second), enough power to translate a million words in less than a second. “Due to the oversupply of renewable energy, we can train our neural networks very cost-effectively in Iceland. We will continue to invest in high-performance hardware there,” explains CTO Jaroslaw Kutylowski.
The DeepL team uses this supercomputer to train neural networks on a huge collection of multilingual texts. The networks look at a lot of translations and learn independently how to translate grammatically correctly and choose good wording. The Cologne-based company can rely on the success of its first product: Linguee, the world's largest search engine for translations. Over the last decade, DeepL has collected more than a billion high-quality translations and can now use them to train artificial intelligence.
The DeepL Translator currently supports 42 language combinations between German, English, French, Spanish, Italian, Polish and Dutch. The neural networks are already training to be able to master other languages such as Mandarin, Japanese and Russian in the future. DeepL also plans to make an application programming interface (API) available in the coming months. The superior translation technology can also be used in other products such as digital assistants, dictionaries, language learning applications and professional translation programs.
“Our ambitions are not limited to translations,” says Frahling. “The neural networks have developed an incredible understanding of language. This opens up many exciting opportunities for us in the future.”
Notes:
1) This content reflects the current state of affairs at the time of publication. The reproduction of individual images, screenshots, embeds or video sequences serves to discuss the topic. 2) Individual contributions were created through the use of machine assistance and were carefully checked by the Mimikama editorial team before publication. ( Reason )