👨🏻💻 Google Summer of Code 2021 @ Apertium.org
2 min readAug 22, 2021
- General Information 📝
Organization: Apertium
Project: Adopting the Hindi-Bengali language pair (unreleased language pair)
Student: Gourab Chakraborty
Mentor: Hèctor Alòs i Font - Abstract 📜
This summer, I was selected for GSoC to work for Apertium, which is a Rule Based Machine Translation platform. It is a really cool organisation. Here we have dictionaries and use rules to make translations. Linguistics is heavily used for making these rules along with other aspects of CS. As a result, the process of making a language pair can take time and a lot of attention to detail. But once it’s created, it gives very accurate results. The primary focus of the orgaisation is to make language pairs for languages which are rare and aren’t supported by Google Translate since it’s not possibe for Google to translate complicated languages by their generalised algorithm. - Goals 🎯
My goal for the summer was to develop the Hindi-Bengali language pair for Apertium. Previously there was no official work done under Apertium for this pair. I started the build on top of a very basic existing opensource project (which inturn used Apertium). - Results 🚀
During the coding period we faced various challenges and overcame them. Now we have a built a foundation of a Hindi-Bengali translator which can be moved towards a more final production build with some more effort from the future developers. - Contributions 🎁
My git commits in the various Apertium repositories are compliled in the following website, thanks to Tino Didriksen
- Communication (Chat) 🗣️
We communicated via IRC channels and over E-mails. - Documentations:
https://wiki.apertium.org/wiki/User:Gourab337
https://wiki.apertium.org/wiki/Hindi_and_Bengali - Future Plans ⏲️
I plan to keep this project alive. There are a few deliverables that I wasn't able to complete due to time constraints. And I plan to complete those and keep further contributing to my language pair. - Key Takeaways 🧠
This summer I learnt a very important lesson in "quality over quantity" and productivity in general. I would never have been able to learn this if not for GSoC, Apertium and my mentor this early on coming from a place where quantity is given a priority. This international open-source experience was very impactful for me. And I'm very thankful to my mentor for giving me this opportunity to work on my language. - Acknowledgement 👏
Apertium is a very active community and very serious work gets done here. There were times when I did errors and made mistakes. But my mentor guided me through the thick and thin. I'm very grateful to him for that. Without him and Apertium, I would have missed out on a very important experience of my life. I would also like to thank Google for funding and promotion open-source among students.
