Locating errors in combinatorial testing using set of possible faulty interactions

In this paper, we present an algorithm for locating errors in combinatorial testing, based on the set of possible faulty interactions (SPFI) constructed according to the result of combinatorial testing. By introducing the error density of each test case or interaction, we can analyze every test case which caused faults of the software under test, and then the additional test cases are generated to locate the faulty interaction. By constructing the set of possible faulty interactions, the same faulty interactions covered by different test case is avoid to be located repeatedly. And all of the faulty interactions can be located exactly when SPFI is empty. The final empirical results show that the number of test cases needed to locate all faulty interactions can be effectively reduced.