Venturing with Similarity-Based Testing: An Approach

Abstract

As well as software systems continuously evolving, test suits are frequently growing. It is impractical to rerun all test cases when the resources are limited, too. Coverage-based test case ranking techniques have been proposed to improve testing effectiveness. In the direction of the path of Test case generation, a method for test case selection with minimal human intervention based on the use of a similarity function is explained. In this context, a similarity-based test case outranking approach has been proposed. Under the proposed technology, a test case generation algorithm using a Labeled Transition diagram is proposed as input. Secondly, the proposed approach measures the path length by designating the similarity matrix to generate reduced executable test cases. This methodology is designed to select less similar test cases and provide the most potential coverage of the functional model from which test cases are generated. An empirical study on two applications has been designed and conducted to formalize the effectiveness of the proposed approach.