MUMBAI, India, Sept. 14 -- Intellectual Property India has published a patent application (202621034578 A) filed by Nilima Dinesh Bobade; Dr. Swati S. Sherekar; and Pranali Ramesh Landge on March 23, 2026, for Optibert Sql Injection Detection System Using Transformer-Cnn Hybrid Model.
Inventors include Nilima Dinesh Bobade; Dr. Swati S. Sherekar; and Pranali Ramesh Landge.
The application for the patent was published on September 11, 2026, under issue no. 37/2026.
Abstract: The present invention discloses a system and method for detecting and preventing SQL Injection attacks in web applications using a hybrid deep learning architecture. The system, referred to as OPTIBERT, integrates transformer-based Natural Language Processing with convolutional neural networks to analyze and classify SQL queries. Input queries are preprocessed through normalization and tokenization, followed by contextual embedding generation using a lightweight transformer model. The embeddings are further processed by a convolutional neural network to extract structural patterns and classified using dense neural layers as benign or malicious. A decision module prevents execution of malicious queries while allowing legitimate queries to proceed to the database. The invention reduces reliance on rule-based detection mechanisms and improves accuracy by capturing both semantic and structural characteristics of queries. The system provides a scalable, efficient, and adaptive solution for protecting web applications against evolving SQL Injection threats.
Disclaimer: Curated by HT Syndication.
