A Comparative study of various Mutants Reduction Techniques

Abstract

Generating huge number of mutants program as possible combinations of various operators, functions, and statements is a great challenge towards the way to success of mutation testing techniques. Programmers and software developers’ tries to escape mutation testing cumbersome and expensive procedure due to this reason. If we succeed to decrease the amount of mutants program without disturbing the quality of mutation testing results then it may prove great technique for software testing as well as popular among programmers. Researchers tries to develop such techniques by which we can decrease the amount of mutants’ programs significant level without affecting the quality of the mutation testing. Among mutants reduction techniques; Sampling method, Clustering Method, Selection Method and Higher Order Mutants are very popular. Rating these 4 available techniques on the basis of their mutants reduction capabilities is still a research problem. In this research we successfully demonstrate and compared the above 4 mutants’ reduction techniques. We find the sampling technique outperformer amongst all that could decrease the amount of mutants up to 90.42% while the cluster method could reduce only 69.42% mutants. The Selection method scored 81.52 % reduction score which is greater than cluster method and lower than sampling method. We also compare the relative performances amongst the mutants’ reduction methods. We conclude that: A) Sampling method is 21% more efficient than Cluster method and 8.9% more efficient than Selection method and ; B) Selection method is 12.1% more efficient than Clustering method. Thus the cluster method is the most expensive and sampling method is most economic amongst the all