Re: Emacs as a translator's tool

[Jean-Christophe Helary <jean.christophe.helary@traduction-libre.org>]

> On Jun 10, 2020, at 20:49, Emanuel Berg via Users list for the GNU Emacs text editor <help-gnu-emacs@gnu.org> wrote:
> 
> Jean-Christophe Helary wrote:
> 
>> Emacs default sentence definition can be expressed
>> as a SRX rules set. SRX is useful for eXchanging
>> information. But the format you use internally
>> is irrelevant.
> 
> The Emacs sentence format and associated functions
> certainly makes sense to use but with tweaked
> settings to get in particular shorter segments
> I think would benefit both searching the DB and
> getting better results.

Sure, but it's not trivial to find "natural" subsegments with the tools at hand.

> For example this is one sentence by the default
> rules:
> 
>   Both of these are library functions that do a lot
>   under the hood, but if they don't meet your needs
>   (or you just want to experiment and learn) you can
>   also use system calls directly.
> 
> But, would you get a good translation suggestion out
> of all of that?

It depends on what's been translated already, I guess.

> Or is it better to simplify in terms of the computer,
> and increase the interactivity/human checking by
> feeding
> 
>   1. Both of these are library functions
>   2. that do a lot under the hood
>   3. but if they don't meet your needs
>   4. or you just want to experiment and learn
>   5. you can also use system calls directly
> 
> ?

I honestly have no idea how complex matching algorithms work to produce subsegment matches.

> Anyway, if there isn't anything let's drop this, what
> remains is the DB of translation suggestions, and the
> algorithm to search and quantify, so e.g., for (1),
> one would get, if one translates into Swedish
> 
>   segment: Both of these are library functions           hit
>   ----------------------------------------------------------
>   suggestion 1 [a]: Båda två är biblioteksfunktioner     90%
>   suggestion 2 [s]: Båda två är bibliotekets funktioner   7%
>   ...               ...                                  ...
>   suggestion n [~]: Den här översättningen suger          1%
> 
> Then one would hit [a] to insert suggestion 1!
> 
> So yes, where do you get the database?

In most of the cases, it's something the translator build from her own translations.

And then, there is the matching algorithm that produces the most relevant match.

-- 
Jean-Christophe Helary @brandelune
http://mac4translators.blogspot.com