Test case prioritization using a Hybrid Chaotic Flower-fruit fly optimization algorithm with multiple objectives

doi:10.1007/s11042-023-16606-0

2023-09-05

PRE

DOI：10.1007/s11042-023-16606-0

原文链接

原文求助

摘要

En 中文

The research aims to resolve the challenges faced in traditional Test Case Prioritization(TCP) techniques and the need to enhance the efficiency of the software testing process.Software development involves test case execution, which tests the changes to the system that requires more resources and time. TCP is the indispensable method, which is established to obtain the goals, such as attaining fast fault detection and a high convergence rate. The code-coverage- methods are the advanced approach employed in the TCP, which is utilized in various prioritization methods to enhance the efficiencyof the method. In this research, the multi-objective- hybrid Chaotic Flower fruit-fly optimization algorithm (Hybrid CFFO) is proposed for the TCP. Instead of using rigid rules or deterministic algorithms, metaheuristic algorithms were created to make intelligent decisions on heuristics and approximations. The foraging behavior of the flower fly and the reproduction characters of the flower are merged which covers a large exploration phase and overcome the local optima. In TCP, being able to quickly explore interesting search space regions and avoid local optima the proposed hybrid Chaotic Flower fruit-fly optimization is essential for producing high-quality solutions. The selection measures used by the optimization algorithm are on the average percentage of combinatorial coverage (APCC) and the Normalized average of the percentage of faults detected (NAPFD), which is designed to develop the multi-objective function. Based on the algorithm, the significance of the test cases is ordered and prioritized. The effectiveness of the proposed methodology is revealed by comparing the proposed multi-objective-based hybrid CFFO method with the conventional methods. From the results obtained, it is proved that the multi-objective-hybrid CFFO method outperforms all the conventional methods in terms of fitness, NAPFD, and APCC. The maximum fitness, NAPFD,and APCC achieved by the proposed methods are 93.1%,92.82%, and 91.1%, respectively for the big fault matrix data.

Keyword:

Test Case Prioritization

Optimization

Multi-objectives

Coverage

Software defect prediction