User:Kantoro/GSOC proposal 2013

From Apertium
Jump to navigation Jump to search

Apertium-kaz-kir: machine translation between Kyrgyz and Kazakh languages

Contact information

Email: kantoro90@gmail.com
Short description: Extending the existing new pair language (kaz-kir). Work around bilingual dictionary, constraint grammars and transfer rules.

Contact information
Name: Kantoro Erkulov
E-mail adress: kantoro90@gmail.com

Other information that may be useful to contact you:
IRC: kantoro #apertium #hfst
Timezone: UTC+06.00

Why is it you are interested in machine translation?

I was born in country where Russian is very popular. We lived in town where majority of people were speaking in Russian. And I was going to Russian school. So starting from my childhood there was an experience to translate Russian tales to my native Kyrgyz language when I was in village on holidays and telling those tales to my relatives. Then I went to high school and learned from there Turkish and English. When I started studying at university I have met the google.translate.com tool. For now I see that there are rule based translating tools. As my major is Computer Science I'm really interested in this field and want to apply my theoretical knowledge into practical project with my native Kyrgyz language.

Why is it that you are interested in the Apertium project?

As I have seen from the home page of Apertium, It is spending a lot of effort on developing language pairs with Turkic family. And as Kyrgyz belongs to this category I'm very happy that I will be helping to my language and participating in such important open-source project.

Which of the published tasks are you interested in? What do you plan to do?

Title

Kyrgyz-Kazakh language pair. Extending existing language pair.


Reasons Why should Google and Apertium sponsor it?

To popularize Turkic family languages they should be participating in some interesting project like in an Apertium as an example. We understand that today the trend of these languages is going down. And native speakers should do everything to change that situation. Google and Apertium in my opinion are the best examples of how this can be achieved.


How it will benefit the society

Kazakh is a little bit similar to Kyrgyz in pronunciation, but when it comes to writings it can be difficult to understand what's written there. Till now there wasn't such tool. And such tool would be very useful in these situations. It can be used not only as translating tool but in education it also can be applied.


Work plan

List your skills and give evidence of your qualifications

Bachelor in Computer Science at IAAU Bishkek, Kyrgyzstan 2008-2014
Languges: Kyrgyz(native), Russian(fluent), English(good), Turkish(good)
Programming: Java, C#, C/C++
Intern Software Developer at "ProDVD LLC Bishkek" - September 2012 - December 2012
Junior Software Developer at SE "Info-System" under MoF of KR - January 2013 - present.
Have implemented the demo version of www.donors.okmot.kg portal (External Aid Management Portal)


Non-GSoC activities

Currently I'm employed as junior software developer at SE "Info-System" under MoF of KR. But I think I will have 35 free hours a week to work on this project.