MUMBAI, India, Jan. 9 -- Intellectual Property India has published a patent application (202511114395 A) filed by Geetanjali Saini; and Vineet Sinha, Gurgaon, Haryana, on Nov. 20, 2025, for 'system and method for semantic file indexing and training-data optimization using an slm-based keyword-buffer architecture.'
Inventor(s) include Geetanjali Saini; and Vineet Sinha.
The application for the patent was published on Jan. 9, under issue no. 02/2026.
According to the abstract released by the Intellectual Property India: "A computer-implemented system is disclosed for high-speed file retrieval and energy-efficient AI dataset preparation using a dual-mode Small Language Model (SLM) parser. The parser operates either automatically or with user intervention to extract multiple semantic and predictive keywords from files. Extracted keywords are temporarily stored in a buffer for deduplication, clustering, ranking, and validation, then transferred into a Semantic-Link Mapping Index Table linking keywords to files. The index enables fast retrieval without file-content crawling. The system also filters irrelevant files before machine-learning and LLM training, reducing computational load, memory usage, and energy requirements. Technical effects include improved retrieval speed, reduced disk operations, more consistent indexing, supporting semantic processing efficiency, and lower energy demands in AI dataset preparation and model training."
Disclaimer: Curated by HT Syndication.
