On Nov 30, 7:48 am, loupiote wrote:
> > Please give some examples.
>
> here are examples:
Merci! Je travail pour le gouvernement de Canada dans un poste
bilangue, et mon dernière evaluation en français m'a apporté un note
de "CCC", qui est le meillier possible pour un deuxième langue. Mais
malgrés ça, il semble que j'ai encore plussiers des choses à
apprendre.
> so this makes good automatic translation impossible unless the machine
> has an understanding of the context.
Assigning gender to inanimate objects is a result of the extremely
strong sex drive in humans. And when you consider the latent context
in all matters sexual, any human author can end up biasing his words,
either conciously or not. Consider your example "amour". Not only does
it depend on natural gender, but also on sexual orientation.
As well, consider the possibility that the author has possible
embedded some gender-based (or sexual orientation based) humour. If
you can figure out a way to decipher humour, you will probably make a
mint with the creation of a program to write jokes.
Unfortunately, Google translate makes absolutely no attempt to
determine semantic context, or even grammatical syntax. It is
completely based on statistical compilations of existing translations.
It would require a major effort to provide the functionality that you
are asking for.
