Utilizing AI in Test Automation to Perform Functional Testing on Web Application

Abstract

Artificial Intelligence is the trend in software development. Unfortunately, Artificial Intelligence algorithms and technologies are still not utilized enough in software testing. Designing Test automation has become the main job for quality engineers and software testers. Mainly, Test Automation is beneficial in reducing manual testing efforts. Utilizing AI in test automation can form a huge benefit in code optimization and test oracle problem. The primary objective of the research was to approve the usability of the Fuzzy Inference System in providing a test oracle for web application functional testing. The secondary objective was to utilize Artificial Intelligence techniques like self-healing for the test Automation using web scraping. Also, to compare the web scraping approach and the Image processing approach in locating the web elements on the websites dynamically. I have addressed the problem by developing Test Automation that verifies the search functionality for a given website. The hypothesis is mainly to check if the Fuzzy Inference System can predict if the search functionality for a given website is working or not. I tested the hypothesis on ten different websites. Then, after I analysed the results, I have found that implementing the Fuzzy Inference System in test automation can form a reasonable solution for the test oracle problem. Furthermore, using the Fuzzy Inference System is as efficient as the manually prepared test oracle that covers all the possible cases for the inputs using if-else statements. Finally, I have demonstrated how web scraping can be utilized to perform self-healing for the test Automation.