1. Controlling Occurrence Frequencies of Parameter Values in Pair-Wise Testing
- Author
-
Satoshi Fujimoto, Tatsuhiro Tsuchiya, and Hideharu Kojima
- Subjects
Artificial Intelligence ,Computer Networks and Communications ,Software testing ,020204 information systems ,0202 electrical engineering, electronic engineering, information engineering ,Pair wise testing ,020207 software engineering ,02 engineering and technology ,Computer Graphics and Computer-Aided Design ,Algorithm ,Software ,Mathematics ,Weighting - Abstract
Pair-wise testing is a widely used strategy of software testing. It requires testing every pair of parameter values at least once. This paper focuses weighting of parameter values for this testing strategy. Weighting is an added feature which allows the tester to prioritize different parameter values by specifying their desired frequency of occurrence in a test suite. This feature is desirable as it allows the tester to have more control over the resulting test suite. However, there has been not much research on weighting: to our knowledge, all existing weighting methods treat weights as a second class requirement and cannot generate a test suite that sufficiently respects the given weights. Aiming to overcome this problem, this paper proposes a weighting method which can be used in combination of any one-test-at-a-time greedy test case generation algorithm. By comparing the parameter value distribution in the current test suite and the ideal one specified by the given weights, the method generates each test case so that the resulting test suite can reflect the weights as accurately as possible. The usefulness of the method is demonstrated through empirical results.
- Published
- 2018