MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202621094615 A) filed by Symbiosis International Deemed University on August 04, 2026, for System And Method For Detecting Fake News Using Unsupervised Anomaly Detection.
Inventors include Dr. Nikhil Mangrulkar; Yesh Raheja; Vaibhav Dhosewan; and Abhishek Chhapparghare.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: ABSTRACT SYSTEM AND METHOD FOR DETECTING FAKE NEWS USING UNSUPERVISED ANOMALY DETECTION The invention presents a system and method for detecting fake news using unsupervised anomaly detection combined with natural language processing, real-time source verification, and AI-driven contextual analysis. The system receives textual input through a user module, preprocesses it via tokenization, stop-word removal, and normalization, and converts the cleaned text into numerical vectors using TF-IDF. An Isolation Forest algorithm classifies these vectors as normal or anomalous without requiring labeled training data. To ensure reliability, a source verification module cross-checks content against trusted databases through external APIs. Additionally, an AI-based analysis module generates contextual explanations using a generative engine. Finally, a result compilation module aggregates outputs to deliver a prediction label, confidence score, and evidence strength, which are displayed to the user. This integrated approach enables efficient, real-time detection of misinformation with transparent explanations, enhancing trust and credibility in digital news consumption. [
Disclaimer: Curated by HT Syndication.
