Natural Language based Test Automation Model for Web Applications

Abstract

Test Automation is a latest technology in the software industry. People favor the test automation over the manual testing due to significant advantages such as, automation is faster than the manual testing, reduction of the human intervention, quality and improve the accuracy. However, the initial cost of the test automation is high. The main reason for this is the lack of technical expertise required for test automation within the industry. This results to organizations have to invest time, money and effort to train the less technical resources. In current, organizations are highly concerned with finding means of resolving this issue. The proposed model addresses this issue by introducing a Natural Language Processing based test automation model for web applications which can be used for non-technical resources with the minimum technical expertise. It allows a user to enter test steps in English. The model is capable of processing the test steps entered by the user and converting them into automation scripts which can be executed using the automation framework that is built. Furthermore, result reports can also be obtained by the execution of generated scripts. Overall accuracy of the new model has been determined to be 84.33 %.