A script that can read every Indian language?
According to a group of IIT-Madras researchers, a single, unified script called Bharati is underway, which once developed will help read every Indian language. From 22 official languages to 1652 dialects in the country, the team aims to reduce script barriers.
The idea was envisioned by V. Srinivasa Chakravarthy, who is a faculty at Computational Neuroscience lab of Department of Biotechnology, three years ago.
Bharati finds its validity through the fact that most Indian languages have a common phonetic structure with vowels and consonants, despite the scripts being different.
Instead of picking a separate character for each consonant, the script incorporates one base character for each consonant family and a modifying character will be used to differentiate between various consonants that fall under the same family, like ka, kha, ga.
To include certain sounds that may be unique to a particular language, like ‘Chillu’ letters in Malayalam, the team has introduced these as special characters in the script, reports Deccan Chronicle.
According to Krishna Bharadwaj M. S., a research associate who is part of the team, the script will help boost domestic tourism, for many tourist places still have signboards in their native languages.
He also added that the script would prove beneficial for dyslexic students, for its implementation of simple characters.
The idea struck the team, when they working on an ambitious project of developing an Optical Character Recognition (OCR) software that could identify various microorganisms.
Currently in its development phase, the team is training an Artificial Neural Network to read the script and has found over 97% accuracy.
They have also developed word games to make the script popular amidst people and will soon launch an application.