Ahp-port: An Analytical Hierarchy Process (AHP) based methodology for Prioritization of Requirements for Test (PORT) in software testing.

摘要

Quality is an important aspect of any product and quality software is no exception. Software testing aims to evaluate software quality and is an integral part of the software development life cycle. Software testers need to carefully plan for testing of software systems due to the constraints of time and money. One issue which software testers face is deciding on the number of testing needs to be performed in this phase. Choosing appropriate test cases is crucial for deriving maximum benefits with the given time and money constraints. Selecting more than the required test cases leads to unwanted expenditure of time and effort while selecting inadequate test cases leads to an increase in risk of defects in application.;This study proposes a system level prioritization methodology for customer requirements based on four factors: requirement volatility, customer priority, implementation complexity and fault proneness of requirements. The Analytical Hierarchy Process (AHP) technique is incorporated in the existing version of the Prioritization of Requirements for Test (PORT) methodology. A ranked list of test cases associated with the prioritized customer requirements is developed and executed with the aim of increasing the rate of fault detection.;The methodology is implemented for two software applications. A Weighted Percentage of Failures Detected (WPFD) is used as a measure of the rate of fault detection of the ranked test suite. For the first case study, the WPFD of the AHP-PORT ranked test suite was found to be 60.94 while that of the randomly ordered test suite was 46.34. For the second case study, the WPFD of the proposed methodology was 62.50 while that of the random ordering was 44.34. A one-sample t-test with a confidence level of 95% was performed to check whether the results obtained were statistically significant. The results show that the rate of fault detection of a given test suite is enhanced when test cases are executed in the descending order of their ranks