Siberian Federal University

Issue

Journal of Siberian Federal University. Engineering & Technologies. 2022 15 (2)

Authors

Legalov, Alexander I.; Zykov, Sergey V.

Contact information

Legalov, Alexander I.: Higher school of economics Moscow, Russian Federation; Zykov, Sergey V. :Higher school of economics Moscow, Russian Federation

Keywords

testing; intelligent systems; machine learning; modular structure; graphical user interface

Abstract

One of the approaches to automating graphical user interface testing is software systems that reproduce the scenario of user interaction with the software under test. The process of playing such a scenario is based on the ability to simulate user actions. The principles underlying the implementation of such a testing system can have a significant impact both on the reliability of the reproduction of test scenarios and on the degree of integration of the system into the software under test or the environment that runs it. The overwhelming majority of tools are not able to provide high reliability and portability of test scripts. At the same time, the problem of ensuring the quality of work and testing various user interfaces, regardless of the platform that launches it, screen size, and input method, is extremely complex, multifaceted, and still not effectively solved. The modular architecture of an intelligent graphical user interface testing system is considered. To achieve greater autonomy of the intelligent system, various implementation options have been proposed. A brief description of intellectualization based on machine learning methods is given. The elements of the novelty of the presented architectural solution are modular structure, cyclic execution of test script commands, and the presence of an intelligent module capable of recognizing elements of the graphical user interface. On the basis of the concept of intellectualization of the graphical user interface testing process, the modular cyclical architecture of an intelligent testing system has been created and the process and features of the interaction of intelligent system modules are described. The concept of the test script command execution cycle is proposed and approaches to ensuring the autonomy of an intelligent system are considered. The proposed system can be used to solve problems of integration testing, using intellectualization based on machine learning methods, an intelligent system with a modular architecture allows you to achieve a high level of testing reliability

Pages

274–282

DOI

10.17516/1999-494X-0390

Paper at repository of SibFU

https://elib.sfu-kras.ru/handle/2311/145441

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).