The Indian government has undertaken significant AI-driven initiatives to promote multilingual education, particularly for tribal languages, in alignment with the National Education Policy 2020. The BharatGen project, anchored at IIT Bombay, is a multimodal Large Language Model focused on developing inclusive AI solutions that support all 22 scheduled languages. This initiative involves a consortium of leading academic institutions including IIT Kanpur, IIT Madras, IIT Hyderabad, IIIT Hyderabad, IIM Indore, and IIT Mandi.
Key technological advancements include the BHASHINI platform under the Digital India Programme, which hosts a repository of over 360 AI-based language models and provides more than 22 specialized language services including Automatic Speech Recognition, Machine Translation, Text-to-Speech, Optical Character Recognition, and Transliteration. The platform contains a massive dataset corpus of 246 million parallel sentence pairs and 3.7 million monolingual text entries, all publicly accessible through BHASHINI or the AIKosh platform. The National Language Translation Mission through BHASHINI is digitizing large volumes of text and speech data across all scheduled languages, including tribal languages such as Bhili and Santhali.
The Central Institute of Indian Languages (CIIL), in collaboration with NCERT, has developed extensive linguistic resources for tribal languages. For Santhali, this includes a speech dataset comprising 50 hours of audio recordings from 60 speakers, and parallel text corpora covering 5,200 sentences each in Santhali and Mundari. CIIL has developed foundational primers in multiple tribal languages including Mundari, Kharia, Ho-Hindi, Kurukh, Korwa (Jharkhand and Chhattisgarh), Bhumij, Santhali-Hindi, and Malto. Additionally, CIIL has launched a 12-week Santhali language course on the SWAYAM platform and hosts diverse Santhali resources on its Bharatavani portal including language learning repositories, encyclopedias, textbooks, and dictionaries.
All datasets and models are publicly accessible through the Medha Bhashika website (https://medha.ciil.org) and other digital platforms, supporting research and development in Indian languages. The initiatives represent a comprehensive approach to preserving and promoting linguistic diversity through advanced AI technologies.