MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641124065 A) filed by Koneru Lakshmaiah Education Foundation on October 16, 2026, for Adaptive Reinforcement Learning-Based Cognitive Assistance System For Personalised Memory Support And Controlled Cognitive Offloading.
Inventors include Ethakota Asish Israel; Nallajalla Supreeth; Kolluru Likith Kumar; Palleti Satya Sai Naga Sasidhar; Boddhapu Hareendra; and S. Shanmugan.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A cognitive assistance system (100) learns for each user whether, when and how intensively to give memory-support reminders. Context, task and interaction information (110) and stored items (120) are encoded into a user-state vector (130). A reinforcement learning engine (140), in one embodiment a deep Q-network with experience replay and a target network, selects no action, send reminder, delay reminder, increase frequency or decrease frequency. A reminder scheduler (150) sets, defers or clears the next reminder accordingly, and an interface (180) delivers it. Explicit and implicit feedback (160) yields a reward that rewards task success and user independence and penalises unnecessary assistance, and the policy is updated (170). In a simulated evaluation, the adaptive policy achieved 79.50 per cent task success against 75.80 per cent without assistance, with 7.60 per cent unnecessary assistance against 98.25 per cent for fixed reminders.
Disclaimer: Curated by HT Syndication.
