Vāṇītyā - an Indic Language Learning Platform
Year: 2025
Type: article
Abstract: Vāṇītyā – An Indic Language Learning Platform presents an integrated approach to language education thatleverages advanced machine learning techniques and modern full-stack development to create an adaptive,personalized learning experience. In addressing real-world challenges, Vāṇītyā specifically targets commondifficulties faced by learners of Indic languages, such as mastering the nuances of phonetics and scriptvariations. For instance, learners often struggle with distinguishing subtle differences in vowel sounds—afundamental yet challenging aspect of language acquisition. By dynamically adjusting the frequency anddifficulty of exercises based on user performance, the platform not only reinforces core language skills but alsobuilds confidence and competence over time. This adaptive, data-driven approach demonstrates a significantadvancement in educational technology, offering a scalable solution to overcome the barriers encountered intraditional language learning environments. The platform synergizes a fine-tuned large language model, trainedon over 1.5 million tokens sourced from IndicTrans, with a custom-built reinforcement learning model togenerate, evaluate, and tailor language exercises in real time. This dual-model architecture supports diverselanguage exercises—including transliteration, translation, and interactive quizzes—and seamlessly adapts toindividual user performance
Authors: Mr. S. Sathish kumar, Mrs. Beulah J Karthikeyan, Akshay Shetty, Om Verma, P. Koushik, Neelam Jayanth
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W6949048093
FWCI:
Cites:
Cited by:
Related to: