An Optimized Surrogate Based Technique for Reducing the Mutation Test Cost Using Satin Bower Bird Algorihm

Abstract

Fault based testing is also known as mutation testing. In this, the faults are voluntarily introduced in the code to check for the feasibility of the program. If the introduced faults does not produce any variations in the program, it results that the program need debugging for perfect operations. It is a costly test due to its intensive operations in finding the bugs. In order to reduce the expense for mutation testing, automated tools and optimization algorithms were introduced. Surrogate based optimization is also a type of reducing mutation testing cost in all possible ways by using three conditions as objective functions. It able to minimize the test case and achieve high mutation score, but only for few code snipped. To bridge this gap, here an enhanced version of surrogate based optimized is proposed. The existing technique is enhanced by using satin bower bird optimization for efficient test case reduction and kill all the mutants in the snippet. The proposed method is tested on few snippets like calculator, hash table and stack and so on using EmuJava. The proposed optimized surrogate based optimization able to outperform the existing technique by having high mutation score of above 80% for all testing Java snippets. It also able to reduce the computational time and cost by reducing the test cases minimum of 950 for stack snippet and only 50 iterations for processing the CGPA snippet. Therefore, the proposed satin bower bird based surrogate optimization able to reduce the computational cost along with high percentage for killing the mutants. Hence, the proposed approach is best for cost minimization in mutation testing.