The translation is produced sounded non-native, primarily as a result of word choice decisions; the engine used words that a native speaker would have been unlikely to in that context. In addition, there are some mistakes, like the difficult, that make the language clearly non-native. The Microsoft translation is the most formal of the translations and this formality – primarily expressed through word choice – contributes to the sense of non-nativeness in this translation. However, while this translation reads like a non-native speaker writing too formally because they don’t know how to write otherwise, formality is appropriate in some cases. Microsoft may be a good engine for more formal settings, such as legal translation or business correspondence. ”
-Samuel
“This engine is able to deliver correctly the main idea czech republic mobile database contained in the source text. However, the choice of words could definitely be improved. Also, the use of verb contraction would make this sentence easier to read (they’ve been instead of they have been). ”
-Alexandra
ModernMT
What our translators say about ModernMT for Spanish to English…
“Long and confusing sentences. The difficult is not relative should be what is difficult is not relative, to better convey the meaning of the source sentence. There is only the difficult is a strange sentence, because it should say there is only what is difficult. I would also change we must only feel empathy to we only have to feel empathy.”
-Patricia
“Like all the engines here, it failed to understand the meaning of the core phrase vivir en un armario which is a version of the English phrase living in the closet, meaning to have not come out as gay, etc. The engine assumes that the speaker is talking about a literal closet and doesn’t recognize in the closet as a set phrase. ModernMT performed fairly well and is at the front of the pack alongside Google and Microsoft. Overall, the translation it produces still sounds non-native, however. The main issues encountered for ModernMT are related to word choice, as the engine makes word choice decisions that a native speaker would generally not make.