MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641089682 A) filed by Koneru Lakshmaiah Education Foundation on July 23, 2026, for Attention-Based Bilstm With Hybrid Optimization For Fake News Detection.
Inventors include K. Varalakshmi; and P. M. Ashok Kumar.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The present invention discloses a computer-implemented system and method for intelligent fake news detection using an Attention-Based Bidirectional Long Short- Term Memory (BiLSTM) network integrated with a Hybrid Whale Multi-Verse Optimization (HWMVO) algorithm. The system preprocesses textual news content, generates semantic embeddings, extracts contextual features using an attention-enabled BiLSTM architecture, and automatically optimizes neural network hyperparameters through hybrid metaheuristic optimization. The optimized model classifies news articles as genuine or fake with improved contextual understanding, convergence, robustness, and computational efficiency. The invention provides a scalable framework for automated misinformation detection across digital communication platforms, social media, and online news environments.
Disclaimer: Curated by HT Syndication.
