MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202611103218 A) filed by Noida International University on August 27, 2026, for Uncertainty-Aware Difficulty-Gated Ontology Expansion Retrieval Framework.
Inventors include Rahul Kumar Singh; and Harsh Vardhan.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: The present invention relates to an uncertainty-aware difficulty-gated ontology expansion retrieval framework for information retrieval systems. The framework generates multiple retrieval rankings including original query ranking, transformer expansion ranking, ontology expansion ranking, and knowledge graph embedding expansion ranking. The system extracts uncertainty features including entropy, score dispersion, top-score gap, inter-ranking agreement, ontology-mapping confidence, and KGE-neighbour stability. Based on these features, the system estimates query difficulty and learns query-specific fusion weights for combining rankings. The framework enables comparison between equal-weight reciprocal rank fusion, Bayesian global weighting, and difficulty-gated weighting approaches. The invention is applicable in search engines, knowledge management systems, digital libraries, and intelligent information retrieval platforms.
Disclaimer: Curated by HT Syndication.
