An Auto-Adapted Method to Generate Pairwise Test Data Set

Abstract

The pairwise test data set generation is one of key issues of combinatorial testing. This paper presents a novel auto-adapted method to generate a pairwise test data set. In this method, all test cases are made at a time, which is called all-tests-at-a-time. Firstly, generate a certain of test data sets; these test data sets have the same of test cases. Secondly, chose the best data set and check whether it satisfy the requirements, if not ,go to next step, else the best is selected and the algorithm is end. Thirdly, update every data set: calculate the repeat number of each test case in a data set, chose two or three test cases according to the repeat number; update the selected test cases relies on main factors of each data set. Moreover, the classic examples are used to illustrate the performance of the proposed method. Compared with the existing algorithms, this paper provides an effective pairwise test suite generation method which updates test cases depend on the data set's coverage not any one independent case; it takes into the relationship of every test case consideration not like the traditional methods which also find the current best case. It can help the data set improve its coverage quickly.