MUMBAI, India, Jan. 9 -- Intellectual Property India has published a patent application (202531128592 A) filed by National Institute Of Technology, Jamshedpur, Jharkhand, on Dec. 18, 2025, for 'system and method for detecting sql injection attacks using dual-embedding autoencoder and latent vector correlation.'
Inventor(s) include Chowdhury, Jayanto Kumar; Yadav, Prof. Dilip Kumar; and P. V. S. S. R., Prof. Chandra Mouli.
The application for the patent was published on Jan. 9, under issue no. 02/2026.
According to the abstract released by the Intellectual Property India: "A system and method are disclosed for detecting SQL injection attacks by leveraging a dual-embedding autoencoder model and latent vector correlation analysis. The invention involves training a neural network autoencoder on legitimate SQL queries, where the encoder incorporates two embedding layers to capture both token-level and character-level (or other granular) features of queries. During operation, incoming SQL queries are encoded into latent vectors, which are then compared against a learned normal profile of query behavior. An anomaly detection mechanism computes the correlation or similarity between the query's latent representation and those of known benign queries. If the query's latent vector deviates beyond a threshold (indicating a likely SQL injection payload), the system flags the query as malicious and can block its execution or raise an alert. This approach enables accurate real-time detection of SQL injection attacks, including new and obfuscated injection patterns, with low false positives, and can be implemented in software, hardware, or hybrid forms to protect database-driven applications."
Disclaimer: Curated by HT Syndication.
