El dj 06 de 03 de 2014 a les 18:05 +0530, en/na Madhulika Mohanty va
escriure:
> 
> Hi,
> 
> GSoC 2014 by Apertium promises to be a great opportunity for learning
> and doing good things. I have gone through the possible project
> proposals. Being a Doctoral student in the field of ML and IR, I would
> love to join you as intern. 
> 
> The project that caught my interest is "Improving support for
> non-standard text input". To begin with, I have installed apertium. I
> read through the requirements of the project. I have read a paper
> titled "A Phrase-based Statistical Model for SMS Text Normalization"
> by AiTi Aw et. al. published in COLING 2006, which talks on similar
> lines. I found this paper to be of relevance for this project. Please
> share your views on the same.  Also suggest some more papers which can
> help me in understanding the problem in depth and writing a better
> proposal. 

That would be interesting, but as Apertium is a rule-based platform,
we'd be looking for any system to be principally rule-based, although of
course ML can be used to learn rules ;)

> Please tell me what all you are looking for in a proposal. Also
> suggest few beginning steps for this project.

The first important step is to install Apertium. If you are a doctoral
student, you should perhaps set your sights on one of the more
challenging tasks. Do you know anything about parsing ? 

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