5 ms·
I just tried the following, to German, and got the predictable result, which is completely wrong. Until the proper semantic parts of the text are identified, th
by nicois 9y ago
I just tried the following, to German, and got the predictable result, which is completely wrong. Until the proper semantic parts of the text are identified, these sorts of mistakes prevent statistical methods from being trustworthy:
Sailing ships in rough seas is asking for trouble
- romseb 9y agoPut a period at the end of the sentence and it changes from > Segelschiffe in rauer See verlangen Ärger. to > Segelschiffe in rauher See sind ein Problem.
- Freak_NL 9y agoThe translating software can't distinguish between 'sailing ships' the noun, and 'sailing' (verb) 'ships' (noun) the act. The latter is what is probably intended. It is also missing chances of using idiomatic expressions in the translations.
- claudius 9y agoWell, 'is' is singular so it can’t be the correct verb associated to 'sailing ships' if the latter means multiple sailing ships rather than the activity of sailing ships.
- ddmd 9y agoI like that I can clikc a word in the target box and choose an alternative translation while the system updates the result accordingly. Previously, I had to play with different translation alternatives manually in Google or Yandex translate. I think deepl also learns from such user choices.
- icebraining 9y agoGoogle Translate lets you do that too..
- akkat 9y agoIt used to, but not anymore for long sentences. That is one big reason that I switched.