[Student-projects] On Spellchecker

[Santhosh Thottingal]

Hi,

I am getting a lot of emails about this project and we saw some of the 
discussions in this mailing list.

This project is complex. I appreciate everybody who read the summary and 
decided to try it. This idea was listed last year too, and we did not 
select anybody. The reason is we did not see a proposal that showed 
understanding of the complexity of the problem.

I expanded the project summary a bit more. Please read it again at 
http://wiki.smc.org.in/SoC/2014/Project_ideas#A_spell_checker_for_Indic_language_that_understands_inflections

I would like to ask everybody interested in this project to spend some 
time to do the homework I listed there. That will help you to understand 
the project.

If you have an algorithm, if you can show some proof of concept, that 
would be great.

For those interested here is a document I wrote many years back on the 
problem http://thottingal.in/documents/MalayalamComputingChallenges.pdf 
- It is bit outdated, but will help you to understand the problem.

If it is a complex project, why listed as a GSOC project?
Ans: First of all, none of our projects are for just finishing at the 
end of the project and forget. So a commendable effort towards  the 
solution is also a best step to begin with. What is important is an 
algorithm and implementation plan. Secondly, I am trying the luck of 
finding bright students and their awesome ideas to solve the problem.


I also want to point out the fact that you need not limit your ideas to 
the list in the wiki. I personally consider student's own ideas more 
valuable than the ideas we listed. So don't be shy to propose your crazy 
project ideas.

All the best!

Thanks
Santhosh