MUMBAI, India, May 1 -- Intellectual Property India has published a patent application (202641026332 A) filed by Dr. T. Vigneswari, Kovilvenni, Tamil Nadu, on March 6, for 'automated test case generation for banking software based on siamese neural network.'
Inventor(s) include Dr. T. Vigneswari; R. Nanthini; R. Sangeetha; U. Sujitha; and G. Yokadharani.
The application for the patent was published on May 1, under issue no. 18/2026.
According to the abstract released by the Intellectual Property India: "This invention proposes a system and method for automated generation of software test cases using a domain-specific artificial intelligence framework. The invention is particularly applicable to banking software systems that involve complex functional modules and require high testing accuracy and traceability. The proposed system processes requirement documents through a pre-processing pipeline that performs normalization, tokenization, and domain vocabulary alignment to extract meaningful representations. A custom Siamese neural network trained entirely from scratch on banking requirement datasets is employed to learn semantic similarity between new requirements and historical requirement-test case pairs. Based on similarity matching, the system automatically generates structured and reusable test cases using intelligent templates that include test scenarios, execution steps, and expected outcomes. The framework operates without reliance on pretrained language models, thereby reducing computational ov~rhead and improving domain adaptability. The invention enhances test coverage, minimizes manual effort, and ensures consistent validation of banking applications. Consequently, it provides a scalable, efficient, and intelligent solution for requirement-driven software testing in regulated environments."
Disclaimer: Curated by HT Syndication.
