An efficient automated test case generation system using gcsru with pattern recognition in software testing

A test case is a detailed set of inputs used to verify the software requirements. In software testing, generating test cases manually is repetitive and time-consuming. Therefore, automated test case generation was developed. The existing works failed to concentrate on non-functional properties of software systems. Consequently, the proposed work presents GCSRU with pattern recognition for test case generation, which covers non-functional properties. First, the collected input data undergoes preprocessing. Next, the functions are extracted, control flows are identified from preprocessed data, while HSC identifies patterns from preprocessed data. Subsequently, AZG-LSOA handles the complexity of the extracted functions and identifies control flows. Following this, BKLI analyses non-deterministic behaviour and generates a state transition matrix from complexity-handled data. Further, features are extracted from the state transition matrix and complexity-handled data. At last, the test case gets generated from GCSRU based on features and patterns and analyzes the non-deterministic behaviour of the system. Hence, the proposed work generates better test cases with 98.36% accuracy than existing ones.