MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202611111769 A) filed by Ims Engineering College, Ghaziabad on September 17, 2026, for A Context-Aware Secure Recommendation System Using Dynamic User Profiling And Edge-Based Preference Adaptation.
Inventors include Dr. Nishant Anand; Dr. Ramesh Kumar Verma; and Mayuri Kulshreshtha.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: [047] The present invention discloses a context-aware secure recommendation system using dynamic user profiling and edge-based preference adaptation. The system comprises a user signal acquisition module, an edge context filtering module, a dynamic profile generation module, a contextual weighting engine, a privacy control module, a preference drift detection module, a recommendation generation engine, a confidence gating module, and a recommendation output interface. The system locally filters raw contextual user signals and generates a compact contextual feature representation to reduce latency, bandwidth consumption, and privacy exposure. A contextual user profile state is generated by combining long-term preference memory with short-term contextual behaviour. The contextual weighting engine dynamically assigns weights to profile attributes based on recency, reliability, frequency, and behavioural consistency. A preference drift detection module identifies temporary or permanent change in user preference, while a confidence gating module controls recommendation output based on contextual reliability. The system thereby provides accurate, privacy-preserving, and technically stable recommendations. Accompanied Drawing [FIGS. 1-2]
Disclaimer: Curated by HT Syndication.
