MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641064143 A) filed by Srm Institute Of Science And Technology Tiruchirappalli on May 21, 2026, for Fake News Detection Using Graph Neural Networks On Propagation Paths.
Inventors include Deepa K; Adhithiyan B; and Dr. Jagadeesh Kannan R.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: A system and method for early fake news detection on Twitter using temporal-aware graph neural networks. The invention analyzes how information propagates rather than just what it says, focusing on timing patterns unique to Twitter misinformation. The system builds time-stamped propagation trees from retweet and reply data, applies a time-weighted graph neural network that prioritizes recent interactions, and performs progressive classification at multiple propagation depths with confidence-based early stopping. Key innovations include bot-aware attention mechanisms targeting coordinated networks and temporal heatmap visualizations showing propagation velocity patterns. The system achieves 85% accuracy with 500ms latency, enabling real-time detection during early spreading phases before viral dissemination.
Disclaimer: Curated by HT Syndication.
