MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202611079523 A) filed by Ajay Kumar Garg Engineering College on June 28, 2026, for Hybrid Multi-Modal Fraud Detection System For Ai-Generated And Collusive Review Identification And Working Method Thereof.
Inventors include Uday Tyagi; Shreyash Chaurasia; Sardul Raajendran; Shubh Gupta; Dr. Anu Chaudhary; Sharma Ji; and Swati Tomar.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The present invention discloses a hybrid multi-modal fraud detection system (100) for identifying AI-generated fraudulent reviews and collusive review activities across digital transaction platforms, e-commerce ecosystems, and computational trust environments. The system comprises a data acquisition and ingestion module (102) configured to collect and preprocess review datasets, a textual intelligence extraction module (104) configured to generate semantic intelligence using TF-IDF and transformer embeddings, a behavioral analytics module (106) configured to extract temporal and interactional fraud indicators, and a dynamic graph intelligence module (108) configured to construct reviewer-entity interaction graphs for collusion detection. A feature fusion and optimization engine (110) generates unified fraud representations processed by a hybrid decision classification engine (112). An explainability and interpretability module (114) generates human-readable fraud reasoning. The system provides improved fraud detection accuracy, reduced computational complexity, enhanced interpretability, and real-time deployment capability on commodity computing infrastructure. Accompanied Drawings [Fig. 1-4]
Disclaimer: Curated by HT Syndication.
