MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202641098918 A) filed by Vellore Institute Of Technology on August 15, 2026, for Towards Robust Security: Multimodal Multi-Layer Defence Against Malicious Urls.
Inventors include Madhu Viswanatham V; M. Sugavanam; Durgadevi U; K Suganya; and V Rajeswari.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: [0029] A multi-modal, multi-layer defense system detects malicious URLs by extracting lexical, host-based, and content features. These features are analyzed independently across dedicated layers: a lexical layer using Random Forest and Gradient Boosting, a host layer employing Support Vector Machine and decision tree models, and a content layer utilizing Convolutional Neural Network and Long Short-Term Memory models. Outputs from these layers are combined via ensemble learning to classify URLs as malicious or benign. This architecture enhances detection accuracy, reduces false positives, and effectively identifies zero-day and obfuscated attacks. [FIG. 1]
Disclaimer: Curated by HT Syndication.
