Language barrier,Recent grad’s thesis project becomes a mission to better represent Urdu in software

By Adam Zewe | Press contact

April 15, 2020

https://seas.harvard.edu/news/2020/04/language-barrier

Zeerak Ahmed-a Pakistan native studying at Harvard-noticed a severe lack of technological support for the Urdu language, prompting him to develop an Urdu keyboard for smartphones. This language(comprised of 39 letters) is built from 21 basic shapes, making traditional keyboard layouts impractical. Ahmed's innovative approach reduced the number of input keys to these foundational 21 shapes, relying on software algorithms to append the necessary dots and symbols to form complete characters. However, this introduced unique algorithmic challenges, as accuracy became critical. Furthermore, Ahmed's journey revealed that most software programming tools are geared towards Latin scripts, necessitating a ground-up development approach for the Urdu language. Beyond just a keyboard, Ahmed expanded his project, "Matnsaz", to provide a programming infrastructure for developing software in non-Latin languages. He has generously shared his research and corpus of Urdu text with the wider developer community. Ahmed's mission isn't just about a single keyboard; he envisions ushering in a new era where languages like Urdu thrive in the digital age.