Apps & Design

[Alan Ward]

Machine translation is a good example of a non-parallelizable
task. We (people) are not parallel machines, and don't think
that way. Instead, our speech is full of cross-references: from
one part of a sentence to another part, between sentences, 
between paragraphs, plus all the external references (e.g. cultural)
you can think of. So to translate a text, you cannot break it up
into little bits, but must treat it as a whole.

It is failing to understand this that makes most translation software
fail miserably.

Best regards,
Alan Ward


> ----------
> De: Kragen Sitaker <kragen at pobox.com>
> A: beowulf at beowulf.org
> Asunto: Re: Apps & Design
> Fecha: divendres, 30 / juny / 2000 20:29

> "Joel" asks:
> > 1.  What kind of work has been done in applying Beowulf to machine 
> > translation?  Would parallelism help when trying to translate several
texts 
> > into several different languages in the shortest possible time?  Can 
> > existing web-based translation systems (Systran, InterTran, &c) be 
> > parallelised?

> Unfortunately, I don't know anything about machine translation, so I
> can't answer the first question; but as for the second question,
> translating multiple documents is an obviously parallelizable task, as
> the results of the translations are independent.  From looking at
> SYSTRAN output, it sort of looks like results of translations of
> invidual phrases in a document are pretty independent when they're
> further than a sentence or two apart.