Využitie umelej inteligencie na automatizované testovanie používateľského rozhrania

Abstract

This bachelor thesis deals with the use of artificial intelligence in automated testing of web application user interfaces. The aim of the thesis was to design and implement a prototype system that combines browser automation using Selenium WebDriver with a large language model. The proposed solution makes it possible to analyze the current state of a web page, identify relevant user interface elements, iteratively generate the next steps of a scenario based on a goal specified in natural language, and after successful execution create a runnable test in the form of a Selenium script written in Java. The thesis also includes an experimental evaluation of the proposed solution on several types of web applications, including public websites and an enterprise application. The experimental results showed that the proposed approach is applicable to both simpler and multi-step scenarios, while its main advantage is the ability to react to the current state of the page. At the same time, it was confirmed that the success of the solution depends on the quality of page analysis, the accuracy of selecting relevant elements, and the level of specificity of the input goal.