Certain investigations on improving the quality of regression testing process by test case prioritization

Abstract

Software system and their environments change continuously. These changes can affect the systems adversely. Hence a software engineer must perform regression testing to ensure the quality of modified system. To improve the cost effectiveness of regression testing, several researchers have proposed various regression testing techniques. Test case prioritization technique is the one of the regression testing techniques. It schedules the test cases to run more test cases earlier so that we can detect faults earlier and provide earlier feedback to the testers. There are two types of test conditions such as unrelated test conditions and related test conditions. In our proposed method, these two test conditions are collected via kernel fuzzy c means clustering procedure. In this, related test conditions are measured for test condition prioritization. Main objective of test case prioritization is to resolve testing environment and arrange them to utilize the maximum possibility to determining error in the source code. The test conditions prioritization takes place by weight optimization process using Modified Artificial Neural Network (MANN) classification algorithms and Whale Optimization Algorithm (WOA). In the second work, new regression test suite prioritization process is used to prioritize the test cases with the goal of maximizing the number of faults that to be found during the execution. At first the input codes are essential to perform test case generation and these test cases are given to the clustering process. In this, clustering can be performed by Optimal Kernel Fuzzy C-Means Clustering (OKFCM). In this research we consider parameters such as Requirement complexity, Requirement Size, Requirement Modification status, Implementation Complexity and Fault Proneness. These requirements are newline