Comparative Analysis of Test Case Prioritization Using Swarm Intelligence Algorithms to Enhance Fault Detection Rate and Execution Time

Abstract

Test Case Prioritization (TCP) is an essential process in software testing that improves fault-detection rates and reduces execution time by determining optimal test ordering. However, traditional manual testing struggles to manage large-scale test suites, while automated methods often demand significant resources. To overcome these challenges, most researchers use metaheuristic algorithms for their efficiency. In particular, Swarm Intelligence (SI), inspired by the behavior of natural systems, has emerged as a powerful tool for solving optimization problems and offers promising solutions for TCP. Therefore, this study aims to compare two promising SI algorithms, namely the Whale Optimization Algorithm (WOA) and the Particle Swarm Optimization Algorithm (PSO), to improve fault detection rate and execution time. Using sample test suites containing 8 test cases with 10 detected faults and 15 test cases with 15 detected faults as datasets. Further, this research analyzes the performance of these algorithms using measures such as Average Percentage of Fault Detection (APFD) and execution time. The results reveal that the algorithms prioritize performance differently, with WOA achieving higher APFD than PSO in two case studies, at 0.813 and 0.789, respectively. In contrast with PSO, it minimizes execution time compared with WOA, achieving 0.026 and 0.103 seconds in two case studies.