Missatge de Xavi Ivars <xavi.iv...@gmail.com> del dia dc., 9 de set. 2020 a
les 13:53:

> Missatge de Tanmai Khanna <khanna.tan...@gmail.com> del dia dc., 9 de
> set. 2020 a les 11:34:
>
>> Hey guys,
>> I'm writing a system demonstration to be submitted at LowResMT 2020 about
>> the recent project that was done as part of GSoC, titled "Translating the
>> internet into low resource languages with Apertium" (Accepting snazzier
>> title suggestions).
>>
>> As part of this demonstration, I want to show some real world examples of
>> how the new system of markup handling will help the translation of webpages
>> and formatted documents - odt, pptx, rtx, etc. To show this effectively, I
>> need to choose 3-4 released language pairs that are sufficiently
>> syntactically divergent that they show the effect of markup reordering in
>> the translation output. As far as I know, spa-cat is one of our most mature
>> pairs, however I'm not sure how syntactically divergent it is. If it is,
>> then I'm happy to be corrected. If your language pair has had issues with
>> webpage translation and those issues are now solved (ish), then some
>> examples would be really helpful.
>>
>>
> Spanish and Catalan are very similar in terms of syntax. We could
> definitely try to get examples of where diverge the most, but those
> examples would need to be completely synthetic.
>
> Markup handling helps, though, in markup handling on different areas: some
> formats where inline tags are common (like ODT), previous
> formatter/deformatter was splitting words where tags appeared, so
> translation of those has improved quite a lot.
>

Spanish and Catalan diverge syntactically in the use of some prepositions.
For example, *de que* (spa) usually becomes *que* (cat).

The last days there has been some oscillation in the translations when a
quotation mark is just in the middle of *de "que*. See:
https://github.com/apertium/apertium-spa-cat/commit/f6f7ee2f560b7ae817ac87c33278a5c4354090dc

But I'm not sure if there has been a real improvement or not. I have to
look at it more carefully.

Jaume
_______________________________________________
Apertium-stuff mailing list
Apertium-stuff@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/apertium-stuff

Reply via email to