Master of Science (MS)
Program or Discipline Name
Nushwan Al-Nakash, PhD.
Software in today’s world is used more and in different ways as well than ever before. From microwaves and vehicles to space rockets and smart cards. Usually, a software programmer goes through a certain process to establish a software that will follow a given specification. Despite the hard work of the programmer, sometimes they make mistakes or sometimes they forget to include all the possibilities of the question for which they are writing the program, which is very humanly in nature. And for those mistakes, a testing unit is always there.
There are numerous techniques of Software Testing, one of which is Boundary Value Analysis. A modified version of Boundary Value Analysis using input parameters with functional dependency is proposed in this work. The idea is derived from the inter dependency of functions among the input parameters. With this modified algorithm, an automated testing tool is created and implemented. This testing tool shows the advantages of the modified algorithm developed over the Functional Tree Approach and reduces a significant amount of test cases that leads to an exhaustive testing. This modified method will test almost every possible required test case increasing the system’s efficiency. This method will be a very good help for any product based company saving a huge amount of money and time.
Generalized BVA generates 5*n number of test cases where n is number of variables while Function Tree method generates the highest of all three that is n*5^(n-1) and the modified approach generates 7*n + k number of test cases where k is the number of mutants killed at each step. So, it shows that the number of test cases in case of modified algorithm is significantly lower than the Function Tree algorithm while almost similar as regular BVA but it covers more functionalities and features
Maheshwari, M. (2017). Boundary Value Analysis for Input Variables with Functional Dependency. Retrieved from https://digitalcommons.harrisburgu.edu/csms_dandt/2