AUTOMATIC GENERATION OF TEST CASES BASED ON MODELS OF SYSTEM BEHAVIOR USING ARTIFICIAL INTELLIGENCE TO IMPROVE THE QUALITY OF SOFTWARE PRODUCTS

Abstract

At various phases of its life cycle, efficient software testing can drastically lower costs and increase the quality of the program. In order to increase the functionality, quality, and dependability of software products, this study investigates the potential for automating the creation of test cases based on system behavior models through the use of artificial intelligence (AI) techniques. The paper looks at important methods for simulating system behavior, such as finite state machines, state diagrams, and transition diagrams, which form the basis for developing test scenarios. The principles of automatic test case generation using machine learning algorithms, neural networks, and heuristic methods are analyzed in detail. A comprehensive approach to test coverage optimization is proposed, allowing for the reduction of redundant tests while improving testing efficiency by adapting scenarios to software changes. Special attention is given to evaluating the effectiveness of the proposed approaches through empirical studies and comparative analysis with traditional testing methods. The research results confirm that AI-based methods significantly reduce test development time, improve defect detection, and ensure deeper code coverage, which is critically important for modern complex and dynamic software systems. The practical significance of this study lies in the potential implementation of the proposed methods in real-world projects to enhance software quality and security while reducing testing and maintenance costs.