Accents matter
Language is more than mere words.
Accents give it life.
When it comes to Sanskrit, the language of gods, it matters a lot.
We would like to create an open source code to convert Sanskrit unaccented words into accented word.
![]()
As you can see the above image shows the accented version.
Our aim is to develop an algorithm and code which would take the unaccented word (पुरोहितं) as input and give accented word (पु॒रोहि॑तं)as output.
Impact
This would be the first of its kind effort in Sanskrit Natural Language Processing.
At the end of the project, we would be able to
- Create a tool which can give accented version of unaccented words.
- Create accented dictionaries out of unaccented database
Future implications
- Though not included in scope of the present work, Sanskrit Text-To-Speech (TTS) convertors can also benifit from the output of this code, because the machine would know that it has to pronounce grave / acute accent properly.
- The framework can be extended for other languages.
Project Time and Budget
Analysis of Sanskrit grammar texts - 300 hours
Preparation of algorithm - 200 hours
Preparation of code - 200 hours
Testing of code - 150 hours
Documentation of code - 150 hours
Total duration - 1000 man hours.
Per hour cost - 10 USD
Total project cost - 10,000 USD.
The Team
1. Dr. Dhaval Patel - Analysis of grammar texts, preparation of algorithm and code
2. Dr. Shivakumari Katuri - Testing and Documentation of the software.
Our previous efforts
We have already collaborated on two similar projects earlier.
1. Sanskrit noun form generator -
https://github.com/drdhaval2785/SanskritSubanta/2. Sanskrit Verb form generator -
https://github.com/drdhaval2785/SanskritVerb/Both of these projects are complete and their outcome is free for all.
Open source
This is the key word in whatever we do.
The code is free for all, whether the use is commercial or non-profit.
Your contribution would keep us doing service of Sanskrit.
