OT: Machine translations

Jan 26, 2013 24 Replies
Saved reading

I did wonder that - I was about to say that there were no links, and no embedded images. However a closer look does reveal a link, so that is probably the explanation.

At least human translation into English by a non-native English speaker usually gets mostly the right interpretation of the Kanji even if the sense and word order is a bit random. This computer one looks like it has been out on the town and still has a very serious opium hangover.

Not laughed at computer translation so much for ages. The ones into common European languages are pretty good these days except when you stray into exotic territory when they leave the odd word in.

The machine translation might make a bit more sense in combination with the original if you knew about 2000 Kanji which is the bare minimum to be even semi-literate. Then with the aid of a thick a Kanji dictionary you could in principle decode it. Minor irritation with this is that to look up an unknown kanji you need to recognise its radical and work out the number of strokes to draw it - that is how the index works.

=DC=87=C9=ED=CC=EE=B3=E4=CE=EF doesn't quite work.

MBQ

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required