MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611090813 A) filed by Noida Institute Of Engineering And Technology Niet on July 26, 2026, for Dynamic Hierarchical Q-Tensor Neural Network System For Multi-Modal Temporal Cognitive Diagnosis.
Inventors include Mr. Farhan Raza Rizvi; and Dr. Karamjeet Kaur.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention discloses a dynamic hierarchical Q-tensor neural network system (100) for multi-modal temporal cognitive diagnosis. The system comprises a multi-modal data acquisition and preprocessing module (110) receiving textual, behavioural, and multimedia learner inputs; a Q-tensor structural encoding module (120) maintaining a three-dimensional Q-tensor spanning item, skill, and cognitive-context dimensions; an adaptive skill-graph neural network module (130) learning skill-to-skill dependencies through graph attention with dynamic edge weights; a temporal mastery tracking module (140) employing a Transformer-based encoder for time-indexed mastery trajectories; a hierarchical feature fusion network (150); a Bayesian uncertainty quantification module (160) producing calibrated confidence intervals; a composite loss optimisation engine (170); a reinforcement-learning-based adaptive Q-tensor refinement module (180); and a real-time diagnostic inference and interpretability engine (190) deployable on
Disclaimer: Curated by HT Syndication.
