El ds 01 de 03 de 2014 a les 17:45 +0530, en/na karan singla va
escriure:
> Hello Sir, 
> The transfer grammar i mentioned is free and is made according to tags
> of LTAG chunker. For use in Apterium, we need to rewrite those
> re-ordering rules again for the use. Also there is an open source
> Bi-lang dictionary available, i guess of size about 150,000 words.
> though it needs some cleaning and may be extended using synonyms from
> WordNet.
> 
> 
> Also it can be a hybrid system, we don't we include the language
> modelling score for re-scoring the output given by the rule based
> system ( number of hypothesis, as we can keep multiple foreign words
> for a same target word).

Or you can use the lexical selection module in Apertium to learn a
source-language model that performs equally as well as a target language
model for selecting the appropriate translation.

http://wiki.apertium.org/wiki/Learning_rules_from_parallel_and_non-parallel_corpora

> I am thinking on the right track ?
> 
> 
> Or Is there a different idea, from which I can start with ? 
> 
> 
> I guess !! some work has already been started on this language pair,
> Can someone brief me on it and guide me from where to start with ?

For this task check out the coding challenge here:

http://wiki.apertium.org/wiki/Ideas_for_Google_Summer_of_Code/Make_a_language_pair_state-of-the-art

Fran



------------------------------------------------------------------------------
Subversion Kills Productivity. Get off Subversion & Make the Move to Perforce.
With Perforce, you get hassle-free workflows. Merge that actually works. 
Faster operations. Version large binaries.  Built-in WAN optimization and the
freedom to use Git, Perforce or both. Make the move to Perforce.
http://pubads.g.doubleclick.net/gampad/clk?id=122218951&iu=/4140/ostg.clktrk
_______________________________________________
Apertium-stuff mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/apertium-stuff

Reply via email to