Open Access Open Access  Restricted Access Subscription Access

Document-Grounded Generative AI for Tamil Explanations of English Engineering Materials: A Semantic Retrieval Approach

Kannan N, Rosenpranav K. S., Mohan Raja V., Jeeva S., Dr. V. Kamalaveni

Abstract


This paper presents a document-grounded educational assistant for engineering learners who prefer Tamil explanations of English study materials. The system extracts PDF text, creates chunks, encodes them with Tamil-ai/tamil-embed-base, retrieves relevant passages using FAISS, and uses Phi-3 Mini for context-based answers. The revised implementation uses the Gemini API for Tamil-language generation. Two generated quiz-answer-review reports—one based on an introductory Artificial Intelligence PDF and one based on a Python Programming PDF—each recorded 7 correct answers out of 10 (70%). These scores describe the quiz takers’ results, not the accuracy of the quiz-generation system. Inspection of the reports identified fragmented questions, PDF extraction artifacts, and generic explanations. The results therefore support the feasibility of the prototype while showing that quiz quality requires further validation.


Full Text:

PDF

References


Abdin, M., Aneja, J., Awadalla, H., Awadallah, A., Awan, A. A., Bach, N., Bahree, A., Bakhtiari, A., Bao, J., Behl, K., Benhaim, A., Bilenko, M., Bjorck, J., Bubeck, S., Cai, M., Cai, Q., Chaudhary, V., Chen, D., Chen, D., ... Zhou, X. (2024). *Phi-3 technical report: A highly capable language model locally on your phone*. arXiv. https://arxiv.org/abs/2404.14219

Douze, M., Guzhva, A., Deng, C., Johnson, J., Szilvasy, G., Mazaré, M. L., Lomeli, M., Hosseini, L., & Jégou, H. (2024). *The Faiss library*. arXiv. https://arxiv.org/abs/2401.08281

Gemini Team. (2023). *Gemini: A family of highly capable multimodal models*. arXiv. https://arxiv.org/abs/2312.11805

Google. (n.d.). *Generate text using the Gemini API*. Google AI for Developers. https://ai.google.dev/gemini-api/docs/text-generation

Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.-T., Rocktäschel, T., Riedel, S., & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive NLP tasks. *Advances in Neural Information Processing Systems, 33*. https://arxiv.org/abs/2005.11401

Reimers, N., & Gurevych, I. (2019). Sentence-BERT: Sentence embeddings using Siamese BERT-networks. In *Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing* (pp. 3982–3992). Association for Computational Linguistics. https://doi.org/10.18653/v1/D19-1410

Tamil-AI. (2026). *Tamil-embed-base* [Sentence embedding model]. Hugging Face. https://huggingface.co/Tamil-ai/tamil-embed-base

UNESCO. (2023). *Guidance for generative AI in education and research*. https://unesdoc.unesco.org/ark:/48223/pf0000386693

Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, Ł., & Polosukhin, I. (2017). Attention is all you need. In *Advances in Neural Information Processing Systems* (Vol. 30). https://papers.nips.cc/paper/7181-attention-is-all-you-need


Refbacks

  • There are currently no refbacks.