El dl 16 de 04 de 2012 a les 14:46 +0200, en/na Per Tunedal va escriure:
> Hi,
> I have read the article too. I noted two interesting features:

No-one has responded to this, so I will -- but I'm not the most adequate
person to respond.

> 1. "... apertium-transfer-tools implements an alignment-template-based
> approach (Och and Ney 2004) to infer structural transfer rules from a
> relatively
> small, sentence-aligned parallel corpus (Sánchez-Martínez and Forcada
> 2009;
> Sánchez-Martínez 2008). The inferred rules can be then edited by a
> linguist, who
> may also add newrules if necessary, or evenmerged with pre-existing
> hand-written
> rules."

This only works with level-1 transfer, so not really applicable to more
distant language pairs.

> 2. "Package apertium-chunks-mixer allows the integration of bilingual
> chunks
> (sub-sentential translation units) obtained by aligning a parallel
> corpus into a translator
> built using Apertium (Sánchez-Martínez et al. 2009)."

Not sure about this, but you can probably use it once you have a working
system.

Fran



------------------------------------------------------------------------------
Better than sec? Nothing is better than sec when it comes to
monitoring Big Data applications. Try Boundary one-second 
resolution app monitoring today. Free.
http://p.sf.net/sfu/Boundary-dev2dev
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff

Reply via email to