Sangita is a natural language toolkit for Indian languages built in Python. Sangita uses Deep Learning for performing basic Natural Language Processing tasks. Currently it provides support for Hindi only.
Vision
- The corpus available for Indian Languages is very limited. Collection of various open corpus, and scraping the web for more linguistic data to promote research and improve the ease of availibility of linguistic data.
- Promoting the use of Deep and Neural learning for usage in otherwise rule-based NLP tasks
- Use of simple functional code for performing complex tasks.
Development leadership
- Natural Languge Processing and Machine Learning: Samriddhi Sinha (IIT Kharagpur, Google Summer of Code '17)
Academic Advisors
- Neha Prabhugaonkar, Principal Data Scientist;Cuddle