MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202621070946 A) filed by Dharmaraj Bajrang Honmore; and Dr. Nageswararao Naik Bhookya on June 08, 2026, for System And Method For Explainable Multimodal Artificial Intelligence-Based Detection And Verification Of Misinformation In Social Media Content.
Inventors include Dharmaraj Bajrang Honmore; and Dr. Nageswararao Naik Bhookya.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The present invention discloses a System and Method for Explainable Multimodal Artificial Intelligence-Based Detection and Verification of Misinformation in Social Media Content. The invention provides an intelligent computational framework for automatically detecting, verifying, and explaining misinformation disseminated through social media platforms and digital communication networks. The system comprises a data acquisition module configured to collect multimodal content including textual information, images, videos, metadata, user interactions, and contextual information from one or more online sources. A preprocessing module structures and normalizes the acquired data, while a multimodal feature extraction module derives textual, visual, and contextual representations using natural language processing, computer vision, and contextual analytics techniques. The extracted features are integrated through a multimodal fusion engine to generate a unified representation for analysis. A misinformation detection engine employs artificial intelligence and deep learning models to identify misleading, manipulated, or false information and generate authenticity classifications. A content verification module validates detected information through evidence retrieval, source credibility assessment, semantic consistency analysis, and comparison with trusted knowledge repositories. An Explainable Artificial Intelligence (XAI) module generates interpretable explanations, feature importance measures, confidence scores, and decision rationales to improve transparency and user trust. A trustworthiness assessment module computes credibility scores based on detection and verification outcomes, while a reporting module generates alerts, verification reports, and decision-support outputs. The invention further incorporates a continuous learning mechanism for adapting to evolving misinformation patterns. The proposed system provides a scalable, accurate, transparent, and real-time solution for misinformation detection and verification across digital communication environments.
Disclaimer: Curated by HT Syndication.
