A Detection-Based Multi-Objective Test Case Selection Algorithm to Improve Time and Efficiency in Regression Testing

Regression testing is carried out to ensure that changes or enhancements are not impacting previous working software. Deciding how much retesting is required after modifications, bug fixes or before product deployments are difficult. Therefore, Test Case Selection (TCS) select the satisfactory subset of modified test cases from already executed test suites. The testing primary concerns in TCS for regression testing are efficiency (i.e., coverage, fault detection ability, redundancy) and time. The first challenge in TCS concerns the efficiency of multi-objective test case selection. The second challenge is to improve the execution time to detect the changes in a test suite, which makes it impractical to use these efficiency measures as a single goal for TCS. To overcome these challenges, there is a need to introduce an efficient detection-based multi-objective framework to improve the Time and efficiency of TCS. A multi-objective advanced and efficient regression test case selection (ARTeCS) framework is devised to improve the time performance and efficiency of a given TCS objective relative to the other TCS approaches. An algorithm to detect the changes in test cases using multiple TCS objectives. This comparison found that the enhanced ARTeCS algorithm improves redundancy efficiency by 44.02%. The selection technique showed ARTeCS improved the modified change detection by 43.00%, whereas the Hybrid Whale Optimization Algorithm (HWOA) stated 23% and ACO showed 33% only for selected test cases. Regarding average for fault detection, ACO scores 21%, HWOA scores 11%, and ARTeCS scores 31.08% with total execution times of 12, 21 and 09 seconds, respectively. In conclusion, the multiple-objective ARTeCS framework with four test suite selection parameters is more efficient than the existing multi-objective selection framework.