Sentence, paragraph and document translation.

[Trickyt57]

Here are 3 ideas related to the translation of entire documents.

1)
If I want to translate an entire document, I have to scroll through it word by word in Pleco. On the other hand, Google translate will translate an entire document. It would be great to add that feature, to see the whole translation in one go. (I know the Google translations from Chinese are pretty crap, but you can still get the gist of what was written.)

2)
Much easier, but also it would be a first on the Internet, would be a tool which takes a sentence, or a paragraph of Chinese characters, and translates it into Pinyin. That would be a great help if I want to give a speech in Chinese, or tell a Chinese joke, but I can't read all the characters that were written for me.

3)
Some years ago there was a project using millions of UN documents which had been officially translated into multiple languages, including Chinese, to create a database of millions of sentences and their translations. Is there any way that Pleco could tap into that to create a sentence or phrase translation tool? I know it would possibly be too big to download, but how about some kind of online access?

 

[mikelove]

Thanks!

1) Tricky unless we use a service like Google's - something of our own would be beyond our abilities I'm afraid.

2) That we actually are thinking about doing, though we need to be a bit more confident in our text segmentation algorithm first.

3) Usually, large corpuses like this require hefty license fees for commercial use, but we can take a look at least.

 

[Trickyt57]

Not sure if this is the link for the 3rd point. You can use it but it seems to get the maximum some kind of password would be needed.

http://untermportal.un.org/search

 

[Ahmad]

I would prefer to see the translation of a whole text side by side instead of a vertical scroll with separate meanings. I admit the the pleco translation of words is accurate and i have tried many translation services. I just hope for a more friendly way of viewing the translated text in order to understand the meaning quickly.

 

[mikelove]

To be honest we'd be more likely to consider it if Google's API was better - right now you can pretty much just plug in the text and get a plain text result, there's no way for us to see which specific Chinese word maps to which specific text on the other side. So the possibilities for us to add useful things on top of that translation result are quite limited.