Intellectual Property India Publishes Patent Application for Multimodal Rag-Based System For Academic Document Understanding And Language Model Training

MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621080927 A) filed by Symbiosis International Deemed University on June 30, 2026, for Multimodal Rag-Based System For Academic Document Understanding And Language Model Training.

Inventors include Aniket Gope; Suruchi Mishra; Dr. Aniket K. Shahade; Dr. Priyanka V. Deshmukh; Dr. Shruti Patil; and Dr. Ketan Kotecha.

The application for the patent was published on August 21, 2026, under issue no. 34/2026.

Abstract: ABSTRACT MULTIMODAL RAG-BASED SYSTEM FOR ACADEMIC DOCUMENT UNDERSTANDING AND LANGUAGE MODEL TRAINING A multimodal retrieval-augmented generation-based system (100) for academic document understanding and language model training is disclosed. The system comprises a document ingestion module (110) for extracting text, images, tables, and equations from PDF documents, an embedding generation module (120) employing a sentence transformer model (122) for generating dense vector embeddings, a FAISS-based vector database storage module (130) for similarity search, a conversational retrieval module (140) with memory-augmented pipeline (142) for multi-turn interactions, a generative response module (150) for producing media-enhanced responses, a media linking module (180) for associating visual content with textual references, a lightweight language model training module (160) enabling custom TinyGPT model training through a graphical interface, and a web-based user interface module (170) providing a unified no-code platform accessible to non-expert users. [

Disclaimer: Curated by HT Syndication.