A classifier-based approach to reduce the generated mutants
Subject Areas : International Journal of Industrial Mathematics
زینب اصغری
1
*
,
Bahman Arasteh Abbasabad
2
,
عباس کوچاری
3
1 - دانشگاه آزاد اسلامی واحد علوم تحقیقات تهران
2 - Computer Engineering, istinye University, Istanbul, Turkey
3 - دانشگاه آزاد اسلامی واحد علوم و تحقیقات تهران
Keywords: software testing , software mutation testing , instruction classification , error propagation , machine learning,
Abstract :
Mutation testing is a powerful technique to evaluate the quality of test suites .The process of generating mutation tests involves generating a large number of test cases, which can be computationally expensive and time-consuming. This study proposes a classifier-based approach to reduce the number of generated mutation tests that involves training a classifier on a set of instruction features to determine which ones are error-prone.The classifier is trained on a dataset of instruction characteristics for identifying the most effective instructions for injecting mutants.Mutation score is calculated to determine the most effective instructions. The study evaluates the effectiveness of the approach through experiments on several open-source projects. The results show that the approach is able to reduce the number of generated mutants while maintaining high mutation score. This approach has the potential to significantly reduce the computational burden of mutation testing and improve the efficiency of software testing.